Metabolic footprint of sepsis and septic shock: A narrative review Authors: Anjali Mishra, Deven Juneja Published: 2026 ABSTRACT Sepsis and septic shock are complex, heterogeneous syndromes associated with high morbidity and mortality despite advances in critical care management. Conventional biomarkers used in clinical practice lack specificity and prognostic accuracy to measure and monitor the dynamic metabolic abnormalities occurring in these conditions. Metabolomics, the comprehensive analysis of low-molecular weight metabolites, are powerful tools to define host metabolic responses to infection and inflammation. This review summarizes current concepts in metabolomic research in sepsis and septic shock, highlighting findings from both experimental models and human studies using nuclear magnetic resonance and mass spectrometry-based platforms. Key metabolic alterations involving energy metabolism, amino acids, lipids, mitochondrial function, and “microbial metabolism” biota-derived metabolites are discussed in relation to disease severity, organ dysfunction, and clinical outcomes. The integration of metabolomics with clinical parameters and other omics technologies holds promise for improving diagnosis, prognostication, and therapeutic stratification in sepsis, paving the way toward precision medicine approaches in critical care. Core Tip: Sepsis remains a cause of significant morbidity and mortality despite recent advances. Moreover, patient’s response to therapy and clinical prognosis is often difficult to accurately predict. Given the heterogeneous and multifactorial nature of disease, broad therapeutic approaches often fail to reduce mortality. Thus, there is a critical need for a better understanding of these complex, multifactorial conditions driven by diverse biological insults and causes. Biomarkers such as procalcitonin, presepsin, interleukin-6, and C-reactive protein are frequently used in diagnosing and monitoring sepsis, but they are far from ideal. Human metabolic activity is extremely sensitive to the surrounding microenvironment, with metabolite profiles reflecting combined influences from transcriptional, translational, and environmental factors. Hence, metabolites are a highly promising class of biomarkers which may be exploited to detect disease presence, progression, and therapeutic response. Emerging data suggests that early introduction of metabolomic data into routine critical care practice may enable more precise severity identification, targeted therapeutic strategies, and improved outcomes for patients with sepsis and septic shock. INTRODUCTION Surviving sepsis campaign has defined sepsis as a life-threatening organ dysfunction caused by a dysregulated host response to infection, and septic shock as its subset where underlying circulatory and cellular/metabolic abnormalities are severe enough that they increase mortality substantially[1]. The Global Burden of Disease Study 2017 estimated that sepsis was responsible for roughly 48.9 million cases and about 11 million deaths globally, accounting for nearly one-fifth of all deaths worldwide. The incidence also varied notably across continents, with a notably higher burden observed in low- and middle-income countries[2]. Despite advances in therapy over recent decades, patient outcomes remain suboptimal and are often difficult to accurately predict. Sepsis arises from complex pathogen-host interactions that frequently dysregulate and exaggerate the body’s initial response to infection[3]. These interactions cause widespread alterations, impairing mitochondrial metabolism and function. In sepsis-associated organ failure, inflammation and oxidative stress cause significant mitochondrial damage, leading to mitochondrial dysfunction and eventual cell death[4]. Although the links between inadequate tissue oxygenation, ischemia-reperfusion injury, hemodynamic instability, inflammation, and progression to multiple organ dysfunction syndrome (MODS) have been extensively studied, the precise molecular mechanisms that ultimately trigger tissue functional injury remain poorly understood[5]. Biomarkers such as procalcitonin, presepsin, interleukin (IL)-6, and C-reactive protein have proven efficacious in diagnosing and monitoring sepsis. However, they often lack sufficient specificity and can be elevated in other inflammatory conditions, such as trauma, burns, and postoperative states[6,7]. Therefore, clinicians globally continue to face a dilemma in accurately diagnosing and managing sepsis with better targeted therapy. Given the heterogeneous and multifactorial nature of disease, broad therapeutic approaches often fail to reduce mortality. Thus, there is a critical need for a better understanding of these complex, multifactorial conditions driven by diverse biological insults and causes[8]. Omics is a newly described term that refers to large-scale study of biological molecules including disciplines like genomics, transcriptomics, proteomics, and metabolomics. Among these, metabolomics, which entails the comprehensive study of all small molecule metabolite profiles to elucidate their impact on the structural, functional, and dynamic aspects of an individual, presents substantial promise in the field of clinical medicine including sepsis and septic shock[8]. DEFINING METABOLOMICS AND ANALYTICAL PLATFORMS Roger Williams first introduced the idea of a “metabolic fingerprint” in 1940, presenting it as a distinctive biochemical pattern unique to each person[9]. The discipline known as “metabolomics” was later developed to systematically explore the metabolome-the collection of small molecules generated during cellular metabolism in living organisms[10]. This area of research encompasses all intracellular chemical processes and acts as a highly effective platform for detecting and precisely quantifying low-molecular-weight metabolites across biological samples[11]. Human metabolic activity is extremely sensitive to the surrounding microenvironment, with metabolite profiles reflecting combined influences from transcriptional, translational, and environmental factors. Because of this, metabolites are a highly promising class of biomarkers for disease presence, progression, and therapeutic response[7,12]. Metabolomics studies have been conducted using two primary strategies: Targeted and untargeted. Targeted metabolomics focuses on the accurate measurement of a carefully selected group of metabolites associated with a specific biochemical pathway or clinical condition. This approach depends on established reference standards; metabolites are extracted from biological samples (such as tissue homogenates, cell cultures, blood, or other fluids) and quantified against previously established calibration curves. This focused, hypothesis-driven technique enables investigators to address highly specific, well-defined biochemical questions[13,14]. Untargeted metabolomics, also known as global metabolomics, employs a comprehensive, hypothesis-free screening strategy that can simultaneously identify thousands of metabolites in a single run. This makes it suitable for exploratory research and for detecting previously unknown metabolites. Biological samples (tissue extracts, cells, or biofluids) are first separated by liquid chromatography technique such as ultra-high performance liquid chromatography (UHPLC), then analyzed using high-resolution mass spectrometry (HRMS) to acquire data[15]. The combination of UHPLC with HRMS provides effective separation with high sensitivity for metabolite detection and quantification. This approach enables coverage of the broad diversity of metabolomes, allowing detection of metabolites present at concentrations even below 0.01%. Systematic identification of these metabolites aids in identification of their metabolic pathways, response to therapeutic interventions and development of various diagnostic and prognostic markers[16]. A range of analytical platforms support metabolomics investigations. Commonly used platforms in sepsis and septic shock studies are nuclear magnetic resonance (NMR) spectroscopy, gas chromatography-mass spectrometry (GC-MS), and liquid chromatography-mass spectrometry (LC-MS), each platform offering distinct advantages and limitations as described in Table 1[13,17]. Table 1Platforms used in metabolomics studies of sepsis and septic shock Analytical platform Metabolite coverage Major advantages Key limitations Relevance in sepsis research ¹H nuclear magnetic resonance Organic acids, amino acids, sugars, selected lipidsHigh reproducibility; minimal sample preparation; quantitative; non-destructiveLower sensitivity; limited detection of low- prevalence metabolitesMetabolic pattern recognition, disease stratification, longitudinal monitoringGas chromatography-mass spectrometryVolatile and derivatized metabolites, organic acids, short-chain fatty acidsExcellent chromatographic resolution; robust compound identificationDerivatization required; limited to volatile/semi-volatile compoundsAssessment of microbial-derived metabolites and energy metabolismLiquid chromatography-mass spectrometryLipids, acylcarnitines, amino acids, bile acids, microbial metabolitesHigh sensitivity; broad metabolite coverage; targeted and untargeted analysisIon suppression; batch effects; requires rigorous standardizationBiomarker discovery, prognostic modeling, metabolic phenotyping Platforms used in metabolomics studies of sepsis and septic shock NMR spectroscopy NMR is a highly reproducible, rapid, non-invasive and quantitative technique. This technique is non- destructive, implying that the sample remains viable and intact for repeat analysis or complementary testing with other methods such as mass spectrometry if later required. NMR permits direct analysis of intact biofluids and tissues without extensive sample preparation or separation, thereby minimizing analytical variability. As a non-destructive technique, it preserves samples for repeat analysis or complementary testing with other methods such as mass spectrometry. It identifies unknown metabolites and supports metabolic pathway tracing through the use of stable isotope-labeled substrates and can detect metabolites across multiple nuclei, including ¹H, ¹³C, ³¹P, and 15N. The methods involved in sample treatment further makes NMR particularly suitable for analyzing labile metabolites (such as glutamine and coenzymes), that are sensitive to ionization conditions used in mass spectrometry. Because of its high reproducibility and minimal sample preparation, is useful for serial monitoring in critically ill patients. It allows repeated assessment of energy metabolism, amino acids, and acid-base-related metabolites, making it suitable for prognostication and follow-up of metabolic trends during the course of sepsis[18,19]. GC-MS GC-MS integrates gas chromatography for compound separation and mass spectrometry for molecular identification. This technique is sensitive to gases and volatile compounds with low boiling points. However, many biological metabolites are non-volatile and have high boiling points, requiring chemical derivatization (chemically modifying polar, non-volatile compounds to increase their volatility, thermal stability, and detectability) before analysis. Derivatization makes this technique expensive and time-consuming, limiting its application[20]. Despite the limitations, GC-MS remains valuable for the analysis of organic acids, fatty acids, and intermediates of mitochondrial metabolism. These metabolites are closely linked to sepsis-related mitochondrial dysfunction and impaired oxidative metabolism, which are key contributors to organ failure and poor outcomes. The technique can be applied in both targeted and untargeted metabolomic studies[21]. LC-MS LC-MS essentially analyzes non-volatile substances and are particularly useful in identifying metabolites with high boiling points, even when present in low concentrations. Unlike GC-MS, LC-MS does not require complex derivatization procedures, making sample preparation simpler and less time-consuming. Currently, the most widely used in sepsis metabolomics due to their high sensitivity and broad metabolite coverage. They enable early differentiation between sepsis and non-infectious inflammatory states, identification of metabolic patterns associated with severity and mortality, and monitoring of metabolic response to interventions such as antibiotics, fluids, and vasopressors[21,22]. Some of the notable limitations of LC-MS includes ion suppression (matrix effect in which co-eluting substances interfere with ionization of the target analytic molecule, resulting in reduced signal intensity in the mass spectrometer) and batch effect (variance in data results introduced by variations in analytical processing conditions between different sets of samples)[23]. Overall, these analytical platforms can potentially establish a direct association between altered metabolism and clinical outcomes in sepsis. By identifying disease-specific metabolic patterns, metabolomics supports early diagnosis, risk stratification, and treatment monitoring, hence improving clinical decision-making in sepsis and septic shock[24]. ANIMAL STUDIES Historically, various techniques have been developed to trigger systemic inflammatory response syndrome (SIRS), sepsis, and septic shock in murine models. Administration of lipopolysaccharide initiates a broad activation of the innate immune response, producing a synchronised cascade that closely mimics gram-negative sepsis observed in human patients[17]. Another technique, the cecal ligation and puncture (CLP) method, remains one of the most frequently utilized models for sepsis research. It effectively replicates bacterial peritonitis and is widely considered the model that most accurately recapitulates human sepsis. Mice undergoing CLP develop polymicrobial infections, exhibit an early hyperdynamic hemodynamic state, and commonly experience acute lung injury (ALI). This model can also generate the typical temporal immune response observed in humans, characterized by an initial proinflammatory surge followed by an immunosuppressive phase[25]. Furthermore, several potential sepsis treatment protocols that failed to demonstrate efficacy in human clinical trials also failed in the CLP model of sepsis, reinforcing the importance of this model in predicting translational outcomes[26]. Animal metabolomics studies consistently demonstrate that sepsis and septic shock are associated with significant changes in energy metabolism, amino acid turnover, lipid handling, and oxidative stress pathways. Using ¹H NMR (Proton Nuclear Magnetic Resonance) spectroscopy, Lin et al[27] and Izquierdo-García et al[28] showed that CLP-induced sepsis alters metabolites related to glycolysis and ketone body metabolism, with more pronounced elevations of lactate, alanine, acetate, acetoacetate, and formate in non-survivors compared with survivors. Lin et al[27] also evaluated distinct serum metabolic patterns, particularly involving energy-related metabolites that differentiated survivors from non-survivors. Overall, the findings of this study support NMR metabonomics as a promising tool for early sepsis prognostication[27]. Mass-spectrometry-based studies further expanded these findings. Using UPLC-Q-TOF-MS-based metabolomics, Liu et al[27] identified distinct plasma metabolic alterations in rat models of burn injury with and without sepsis. Nine metabolites linked to oxidative stress and tissue injury, including hypoxanthine, proline, uracil, indoxyl sulfate, nitrotyrosine, uric acid, glucuronic acid, gluconic acid, and trihydroxy cholanoic acid, emerged as potential markers to help differentiate septic from non-septic burn states[29]. Subsequent studies also reported widespread disruptions in amino acid, fatty acid, choline metabolism, and acylcarnitine metabolism in CLP sepsis, reflecting mitochondrial dysfunction and altered lipid utilization[30,31]. Complementary endotoxin and translational models have also consistently identified changes in acylcarnitines, citrulline, kynurenine, bile acids, and tricarboxylic acid (TCA) cycle intermediates that distinguished septic states, outcomes, and survival, reinforcing metabolomics as a powerful tool for mechanistic insight and prognostication in sepsis and septic shock[32,33]. The knowledge derived from these animal studies have been subsequently applied in human subjects, with promising results. HUMAN STUDIES Some of the early studies that highlighted the prognostic value of metabolic profiling using NMR spectroscopy were conducted in critically ill trauma patients. Mao et al[34] demonstrated that uninfected SIRS was associated with raised branched-chain amino acids and glucose levels, whereas progression to multi-organ failure was marked by an increase in free fatty acids, creatinine, and lactate. Similarly, Cohen et al[35] reported higher levels of lipids, glucose, ketone bodies, and lactate in non-survivors of septic shock. Collectively, these studies supported lipid dysregulation and lactate accumulation as key metabolic parameters linked to poor outcomes in critically ill patients. In another study, the authors applied NMR-based metabolomics to patients with sepsis-associated ALI and identified four metabolites that were markedly different from those in healthy controls. Metabolites linked to oxidative stress, impaired energy balance, and apoptosis-including total glutathione, adenosine, and phosphatidylserine-were elevated, while sphingomyelin, related to endothelial barrier integrity, was reduced[36]. Schmerler et al[37] applied a targeted LC=MS metabolomics approach to distinguish sepsis from non-infectious SIRS by profiling multiple metabolite classes. They identified acylcarnitine C10:1 and glycerophospholipid PCaaC32:0 as key discriminators of sepsis. In 2013, Langley et al[38] conducted a landmark study identifying metabolic markers associated with sepsis mortality, demonstrating that acylcarnitines effectively differentiated survivors from non-survivors. Sepsis survivors also showed reduced levels of citrate, malate, amino acids, and carnitine esters, along with increased acetaminophen metabolites, compared with patients with non-infectious SIRS[38]. Table 2 summarizes key clinical studies in humans that highlight the role of metabolomics in identifying specific metabolites in sepsis to distinguish it from SIRS, assess disease severity, and predict patient outcomes[34-49]. Table 2Clinical metabolomics studies in sepsis and septic shock Ref. Clinical context Analytical platform Key metabolic pathways Main clinical insight Pandey et al[39] (2023)Treatment response in septic shock¹H NMRKetone bodies, amino acids, choline metabolismResponsive patients- higher choline and glutamate. Lactate, 3 Hydroxybutyrate, and phenylalanine were lowerLi et al[40] (2023)Sepsis vs healthy controlsLC-MS/MSPhenylalanine and tryptophan metabolismIdentified novel aromatic metabolites linked to sepsisFeng et al[41] (2022)Trauma (non-SIRS) vs sepsisLC-MS/MSNucleotide and lipid metabolismNine discriminatory metabolites identified that predicted septic conversionChen et al[42] (2022)Sepsis vs healthy controlLC-MS/MSMulti-pathway disruptionBroad metabolite panel (73 metabolites) predicting sepsis onsetPandey et al[43] (2021)Septic shock profiling¹H NMREnergy metabolism, ketone bodiesDistinct metabolic markers that upregulated vs those that down regulated in septic shockJaurila et al[44] (2020)Sepsis mortality¹H NMRCentral carbon metabolismElevated lactate and citrate linked to mortalityChung et al[45] (2019) Septic shock prognosisUHPLC-MSCarnitine metabolismHigh acetylcarnitine associated with non survivalHuang et al[46] (2019)Severe infection risk stratificationLC-MS/MSAmino acid catabolismPhenylalanine and leucine predicted outcome and prognostication for severe infectionsCambiaghi et al[47] (2018) Septic shock lipidomicsLC-MS/MSPhospholipid remodellingLipidome alterations linked to mortalityLiu et al[48] (2019) Septic shock prognosisLC-MS/MSBCAA and carnitine pathwaysMetabolites (43 key metabolites and 6 Primary discriminators) discriminated survivorsNeugebauer et al[49] (2016)Sepsis vs SIRSLC-MS/MSLipid remodellingDistinct lipid markers (Acylcarnitines, glycerophospholipids and sphingolipids) separated sepsis from SIRSLangley et al[38] (2013) Sepsis survivor vs non-survivor vs non infected SIRSLC-MS/MSMitochondrial and nucleotide metabolismDistinct survivor vs non-survivor metabolic profileSchmerler et al[37] (2012) Sepsis vs noninfectious SIRSLC-MS/MSLipid metabolism (acylcarnitines and glycerophospholipids)Acylcarnitine and glycerophospholipid significantly differed between sepsis and SIRSStringer et al[36] (2011) Sepsis-induced ALI¹H NMROxidative stress and apoptosisMetabolites reflected lung injury mechanismsCohen et al[35] (2010)Trauma-related septic shock¹H NMRLipid and glucose metabolismNon-survivors showed lipid accumulationMao et al[34] (2009)Trauma with SIRS or MODS¹H NMREnergy and lipid metabolismDistinct metabolic profiles in MODS vs SIRSUHPLC-MS: Ultra-high performance liquid chromatography-mass spectrometry; LC-MS: Liquid chromatography-mass spectrometry; MS: Mass spectrometry; BCAA: Branched chained amino acids; ALI: Acute lung injury; NMR: Nuclear magnetic resonance; MODS: Multiple organ dysfunction syndrome; SIRS: Systemic inflammatory response syndrome. Clinical metabolomics studies in sepsis and septic shock UHPLC-MS: Ultra-high performance liquid chromatography-mass spectrometry; LC-MS: Liquid chromatography-mass spectrometry; MS: Mass spectrometry; BCAA: Branched chained amino acids; ALI: Acute lung injury; NMR: Nuclear magnetic resonance; MODS: Multiple organ dysfunction syndrome; SIRS: Systemic inflammatory response syndrome. ROLE OF MICROBIOTA AND MICROBIAL METABOLISM Alterations in the gut microbiome play a critical role in modulating the host response and clinical outcomes in sepsis. Disruption of the normal intestinal microbial balance impairs mucosal immunity and barrier function, thereby increasing susceptibility to infection and facilitating systemic inflammation. The composition and burden of the gut microbiota may have a direct influence the severity of organ injury during sepsis, with evidence indicating that reducing the pathogenic bacterial load reduces inflammatory responses and limits organ dysfunction. Selective digestive decontamination is one such therapeutic strategy that employs targeted antimicrobial agents to suppress overgrowth of potentially pathogenic organisms in the gastrointestinal tract while preserving commensal flora[50]. Sepsis is also associated with profound alterations in the respiratory microbiome, particularly in critically ill patients. Translocation of oropharyngeal microorganisms into the lower respiratory tract is common during critical illness, and the proliferation of potentially pathogenic bacteria in the upper airway significantly increases the risk of lung infection[51]. In sepsis and acute respiratory distress syndrome, surfactant dysfunction, reduced mucociliary clearance, and alveolar oedema create hypoxic microenvironments that favour bacterial survival and growth. Together, these factors promote increased microbial migration from the gut and upper airway to the lungs, reduced microbial clearance, and multiplication of pathogenic organisms within the lungs[52]. Integrating metabolomics with microbiome research offers an important opportunity to advance personalized approaches to sepsis by linking host metabolism to microbial influences. Growing evidence shows close interaction between the host and gut microbiota, affecting metabolic and regulatory pathways during critical illness. Understanding microbial metabolic activity alongside host metabolism is therefore essential for deciphering sepsis mechanisms[53]. Many microbial metabolites are essential for normal organ function. For instance, faecal short-chain fatty acids (SCFA) serve as a primary energy source for enterocytes and support the intestinal immune barrier by limiting bacterial translocation. They are typically utilized within the gut mucosa and, therefore, may not be detectable in the circulation of healthy individuals. Normally, physiological mechanisms tightly regulate circulating microbial metabolites to maintain metabolic homeostasis. Excess microbial products entering the bloodstream are detoxified in the liver through conjugation reactions and subsequently eliminated as water-soluble compounds via the kidneys[54]. Mass spectrometry-based metabolomic studies in animals have demonstrated that many circulating low-molecular weight metabolites are derived from gut microbial activity. For instance, phenolic metabolites such as phenyl sulfate, p-cresol sulfate, phenylpropionylglycine, and cinnamoylglycine are detected in much higher concentrations in conventionally colonized animals as compared to germ-free animals[55]. Levels of phenolic acid, a compound produced by microbial metabolism, have also been found to rise in certain critically ill patients. In large amounts, it may disturb the respiratory chain of mitochondria, altering the functions of organs and tissues, causing MODS[53]. Sepsis-associated aromatic microbial metabolites (AMM) exhibit a correlation with bacterial load and the severity of infection. On comparing patients with varying severity of illness in bacterial infection, it was deciphered that in patients with septicemia, serum concentrations of hydrophilic AMM were markedly higher than in patients with only localized infections of skin and soft tissues. It has also been observed that AMM level directly reflects the severity of the infection and correlates with the presence of inflammation, and higher acute physiology and chronic health evaluation (APACHE) II and sequential organ failure assessment (SOFA) scores. Thus, the microbial load in sepsis is directly associated with a proportional rise in microbial metabolites. This further highlight that sepsis is strongly influenced by bi-directional interactions between the host microbiome. Microbial metabolites, particularly aromatic compounds and SCFA, reflect microbial burden and metabolic activity, correlate with disease severity, and may directly impair cellular and mitochondrial function[56]. Combining metabolomic profiling with microbiome analysis, therefore, provides deeper insight into sepsis progression and therapeutic targeting. CLINICAL APPLICATION OF METABOLOMIC MARKERS Based on the current literature, several potential clinical applications of metabolomics have been proposed for the management of sepsis and septic shock (Figure 1). Potential clinical applications of metabolomics in management of sepsis. Prognostication and mortality predication In sepsis, the early phase of acute inflammatory activation is marked by a shift toward accelerated glycolysis and increased amino acid catabolism. Recent human metabolomic studies have confirmed significant elevations in lactate and alterations in amino acid pathways, including branched-chain amino acids and alanine, consistent with stress-induced metabolic responses during the hyperinflammatory stages of infection. Study by Jennaro et al[57], identified a distinct early metabolic and inflammatory profile in 28-day non-survivors with septic shock, characterized by persistently elevated acylcarnitines and protein biomarkers of inflammation and endothelial activation. Non-survivors showed sustained increases in IL-8 and acylcarnitines over the first 48 hours, along with a slower decline in pyruvate, IL-6, tumor necrosis factor-alpha, and angiopoietin-2 compared with survivors[57]. In the later stages of sepsis, this metabolic profile shows a transitions in the cellular metabolism from the glycolytic pathways toward increased lactate production, disruptions in TCA cycle, and reduced mitochondrial oxidative phosphorylation, contributing to organ dysfunction and poor outcomes[58]. A comprehensive review of 27 studies evaluating biomarker panels for predicting mortality in sepsis demonstrated that the majority of metabolic markers associated with survival outcomes were lipids and lipid-like compounds, as well as organic acids and their derivatives. Among these, the metabolites that displayed the most statistically significant difference between sepsis survivors and non-survivors were lactate and acetylcarnitine, followed by phenylalanine, isoleucine, urea, glutamine, and kynurenine[59]. Several other studies that have examined metabolic predictors of mortality in sepsis and septic shock, show that non-survivors commonly exhibit elevated amino acid and ketone concentrations with reduced levels of fatty acid-derived metabolites[60]. Liu et al[48] compared serum metabolomic profiles of septic shock survivors and non-survivors using samples obtained at admission and at 24 hours. Non-survivors showed higher creatinine, energy-related metabolites, and amino acids, along with reduced glycoprotein levels over time, and key discriminatory metabolites included alanine, glutamate, lactate, pyruvate, N-acetyl glycoprotein, and citrate. These findings also demonstrate that early metabolite monitoring may help assess treatment response in septic shock. Novel biomarkers Acute insults such as sepsis trigger rapid and highly dynamic alterations in both the proteome and metabolome, many of which can be effectively analyzed using MS based approaches. These techniques are therefore widely employed for biomarker discovery, phenotypic characterization, and the identification of novel therapeutic targets[61]. A study evaluated NMR-based metabolomics to assess lactate dehydrogenase (LDH) and phenylalanine hydroxylase activity in patients with sepsis, septic shock, and disease controls by analyzing pyruvate/Lactate (Pyr/Lac) and phenylalanine/tyrosine (Phe/Tyr) ratios. Pyr/Lac showed a negative correlation, while Phe/Tyr showed a positive correlation with APACHE II scores, and both ratios demonstrated strong discriminatory performance on area under the receiver operating characteristic curve, analysis. These findings indicate an increased LDH activity and reduced phenylalanine hydroxylase activity in septic shock[62]. Studies have also reported several surrogate metabolic indicators of inflammation, including an increased kynurenine-to-tryptophan ratio and significant disturbances in lipid metabolism. The kynurenine pathway which begins with tryptophan breakdown and contributes to energy production, is upregulated during immune activation. Elevated kynurenine levels reflect enhanced tryptophan degradation driven by interferon-gamma release from activated T cells. Since tryptophan metabolism is closely linked to immune responses, these metabolic patterns may also serve as potential biomarkers for sepsis diagnosis[63]. Identification of host-pathogen interactions Metabolomics, that give valuable insights into the biochemical changes of cellular phenotypes, have emerged as an important tool for identifying host-pathogen interactions and studying newer therapeutic targets[64]. The host-pathogen interactions across the microbial life cycle, including cellular entry, intracellular replication, transmission, and host defence responses, are characterized by multiple metabolic alterations that can be identified through changes in key proteins and metabolites. Advances in mass spectrometry based techniques therefore enable a detailed molecular characterization of infection from the perspective of both host and pathogen[65]. Abuawad et al[64], demonstrated that metabolomics can reveal how host immune cells undergo specific metabolic reprogramming in response to different bacterial pathogens. The authors examined the effects of secretomes from a gram-positive bacterium (Staphylococcus aureus SH1000) and a gram-negative bacterium (Pseudomonas aeruginosa PAO1) on THP-1-derived human macrophages. While some metabolic alterations were shared, distinct pathogen-specific signatures were identified. Staphylococcus aureus exposure was associated with accumulation of asparagine and L-formylkynurenine and depletion of glycine-related metabolites, whereas P. aeruginosa altered creatine levels and reduced asparagine and L-kynurenine. By profiling pathogen-induced changes in key metabolic pathways, the study illustrated the role of metabolomics in providing insights on host-pathogen interactions, highlighting both shared and pathogen-specific metabolic markers. Assessment of therapeutic response There are limited studies that have monitored therapeutic response to treatment in sepsis and septic shock. A sub-study of a randomized controlled trial evaluating L-carnitine therapy in severe sepsis demonstrated time-dependent differences in 3 metabolites (3-hydroxybutyrate, acetoacetate, and 3-hydroxyisovalerate) between survivors and non-survivors following treatment. Patients who responded favorably exhibited lower baseline levels of carnitine and acetylcarnitine, along with post-treatment increases in methionine, lysine, phenylalanine, and tyrosine. Moreover, individuals with lower ketone concentrations who received L-carnitine experienced more rapid shock resolution and showed a trend toward improved long-term survival[66]. Based on SOFA score, Cambiaghi et al[67] stratified patients with septic shock into responders and nonresponders. Compared to non-responders, responders exhibited a greater decline in myristic and oleic acid levels and a smaller reduction in creatinine. Longitudinal analysis further showed an increase in kynurenine concentrations in responders, a pattern not observed in non-responders. LIMITATIONS Despite growing evidence supporting the utility of metabolomics in sepsis and septic shock, several limitations hinder its routine clinical application. Most studies are limited by small sample sizes, single-center designs, heterogeneity in patient populations, timing of sample collection, and analytical platforms, which reduce reproducibility and external validity[12]. In addition, sepsis-associated metabolic changes are dynamic and influenced by confounders such as comorbidities, nutritional status, organ dysfunction, and therapeutic interventions, making it unlikely that a single metabolite can serve as a universal biomarker[68]. Finally, the limited availability, high cost, and absence of standardization may hinder its widespread applicability. FUTURE PERSPECTIVES Future research should focus on large, multi-center, longitudinal studies with standardized sampling protocols and research methods to improve comparability across studies. Integration of metabolomics with other omics approaches, clinical scoring systems, and machine learning models may enhance diagnostic precision, prognostic stratification, and treatment monitoring. Ultimately, translating metabolomic panels into rapid, bedside-compatible assays will be essential for their inclusion into precision medicine strategies for sepsis and septic shock[69]. CONCLUSION Metabolomic studies have substantially advanced the understanding of sepsis and septic shock by identifying significant alterations in host energy metabolism, lipid handling, amino acid turnover, and host-microbiome interactions. Evidence from animal models and clinical studies demonstrates that metabolic profiling can differentiate sepsis from non-infectious inflammatory states, identify patients at risk of poor outcomes, and provide early insight into organ dysfunction and mortality. Although significant challenges remain, particularly related to heterogeneity, standardization, and clinical translation, ongoing advances in analytical platform techniques, bioinformatics, and integrative multi-omics approaches are likely to enhance the clinical utility of metabolomics. Early introduction of metabolomic data into routine critical care practice may enable more precise severity identification, targeted therapeutic strategies, and improved outcomes for patients with sepsis and septic shock.