The infant initially received conventional mechanical ventilation and empirical ampicillin plus ceftazidime after an early episode of Escherichia coli bacteremia. On day 41 of life, while enteral feeding had been advanced to 60 mL/kg/day, he developed pulmonary hemorrhage and disseminated intravascular coagulation with scattered ecchymoses. His white blood cell count reached 2.5 × 10⁹/L, platelet count 29 × 10⁹/L and hemoglobin level 89 g/L. Blood metagenomic next-generation sequencing detected K. pneumoniae with 10 reads, and a rectal swab was positive for carbapenem-resistant K. pneumoniae.

Clinicians changed treatment to polymyxin B, administered with a loading dose of 25,000 IU/kg, equivalent to 2.5 mg/kg, followed by 15,000 IU/kg, or 1.5 mg/kg, every 12 hours, together with aztreonam at 30 mg/kg every 8 hours. Procalcitonin declined to a nadir of 0.25 ng/mL, but the pancytopenia did not improve substantially. On day 56, at a weight of 1.41 kg, the infant developed progressive abdominal distension, gastric retention, fever up to 37.7 °C, abdominal tenderness and absent bowel sounds, with fresh blood detected on digital rectal examination. His white blood cell count was 1.2 × 10⁹/L, C-reactive protein was 31.8 mg/L and procalcitonin was 7.13 ng/mL.

Imaging showed ascites and pneumatosis intestinalis. Exploratory laparotomy confirmed neonatal necrotizing enterocolitis with appendiceal perforation. Surgeons performed partial resection of necrotic bowel, appendectomy and terminal ileostomy. Postoperative blood mNGS detected K. pneumoniae with high confidence at 3,087 species-specific reads, while ascitic-fluid mNGS detected the organism at 259,429 reads. Ascitic-fluid testing also identified the NDM β-lactamase gene with 101 reads, SHV with 40 reads, APH(3′)-Ia with three reads and tet(A) with three reads. Enterococcus faecium was detected with moderate confidence at nine reads, although blood and ascitic-fluid cultures were negative.

After surgery, the infant received vancomycin at 15 mg/kg every 8 hours in combination with polymyxin B and aztreonam. Low-molecular-weight heparin was also used for thromboprophylaxis under an institutional protocol for critically ill neonates with infectious peritonitis. By day 60, erythema, swelling and progressive dark discoloration had developed over the anterior chest, abdominal wall and right lower limb. The right leg became indurated and tense while the dorsalis pedis artery remained palpable. With C-reactive protein at 13.3 mg/L and persistently low platelets, the lesions raised concern for sepsis-associated skin involvement and possible necrotizing fasciitis. Necrotizing fasciitis remained a suspected diagnosis because it was not confirmed surgically and the lesions resolved with antibiotic therapy alone.

Following multidisciplinary review, a risk-benefit assessment and guardian consent, the clinicians replaced aztreonam with off-label aztreonam–avibactam at 40 mg/kg every 8 hours, comprising aztreonam at 30 mg/kg and avibactam at 10 mg/kg, with each dose infused over three hours. Linezolid at 10 mg/kg every eight hours was added for Gram-positive coverage. On day 64 of life—the fourth day of aztreonam–avibactam therapy—the infant showed marked improvement, including substantial resolution of ecchymoses and abdominal distension, recovery of bowel sounds and a decline in C-reactive protein to 3.7 mg/L.

On day 66, the sixth day of treatment, blood mNGS was negative, linezolid was discontinued and enteral feeding was restarted. By day 70, the tenth treatment day, the ecchymoses and right-leg induration had completely resolved, enteral nutrition was advancing smoothly and C-reactive protein had fallen to 0.6 mg/L. Aztreonam–avibactam was continued for 24 days through day 84 because of the complicated intra-abdominal infection. No dose adjustment was required, and the clinicians identified no adverse events considered attributable to aztreonam–avibactam. The infant transitioned to non-invasive ventilation on day 84, reached full enteral feeding with stable weight gain by day 89 and was discharged after 117 days of hospitalization weighing 2,320 g.

Aztreonam–avibactam combines aztreonam, a monobactam stable against metallo-β-lactamase hydrolysis, with avibactam, which inhibits class A, C and some class D serine β-lactamases and can protect aztreonam from degradation by co-produced enzymes. The report states that the combination was approved in China on 30 June 2025 for complicated intra-abdominal infections and hospital-acquired pneumonia, including ventilator-associated pneumonia, in adults with limited or no alternative options, but had no approved neonatal indication. A phase IIa pediatric trial involving neonates and infants younger than nine months had been initiated in 2024 to provide pharmacokinetic and safety data.

The case has important limitations. The neonatal dose was selected through multidisciplinary consultation and extrapolation from standard neonatal aztreonam dosing, without systematic pharmacokinetic assessment or therapeutic drug monitoring, so optimal pharmacokinetic and pharmacodynamic targets were not established. Pathogen identification relied primarily on mNGS because cultures were negative, and standardized surveillance after treatment withdrawal was not performed to confirm durable pathogen clearance. Concomitant vancomycin, linezolid and surgery also limit attribution of the improvement to aztreonam–avibactam alone. As a retrospective single-patient report without a comparison group, the findings document an observed clinical response and apparent tolerability in one extremely preterm infant, but do not establish the regimen’s efficacy or support broad neonatal recommendations. Prospective pharmacokinetic, safety and efficacy data are needed to define its role in critically ill neonates.

Subject of Research: Off-label aztreonam–avibactam salvage treatment of NDM-producing Klebsiella pneumoniae in an extremely preterm infant

Article Title: From sepsis to necrotizing enterocolitis with appendiceal perforation: a case report of successful salvage therapy with aztreonam–avibactam in an extremely preterm infant with NDM-producing Klebsiella pneumoniae

Article References: He, Y., Huang, J., Du, B., Chen, Z., Huang, L., Shi, J., Zhang, M., Zhang, L., & Tang, J. (2026). From sepsis to necrotizing enterocolitis with appendiceal perforation: a case report of successful salvage therapy with aztreonam–avibactam in an extremely preterm infant with NDM-producing Klebsiella pneumoniae. Frontiers in Pharmacology, 17, Article 1919037. https://doi.org/10.3389/fphar.2026.1919037

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1919037

Keywords: aztreonam–avibactam, NDM-producing Klebsiella pneumoniae, extremely preterm infant, salvage therapy, neonatal sepsis, necrotizing enterocolitis, appendiceal perforation, metagenomic sequencing, case report, surgical source control, sepsis, necrotizing

Sepsis is a systemic syndrome driven by infection and a profound failure of immune regulation. The source estimates that it affects about 49 million people annually and contributes to roughly one in five deaths worldwide. Its pathology is not adequately described as persistent inflammation alone. Early disease may feature intense cytokine and reactive oxygen species signaling, endothelial injury and impaired microcirculation, while some patients later develop immune exhaustion, impaired antigen presentation and susceptibility to secondary infection. These states can overlap rather than following a rigid sequence, creating a difficult therapeutic problem: broadly suppressing inflammation may worsen immune paralysis, whereas insufficient anti-inflammatory treatment may leave organ damage unchecked.

Current management combines antibiotics with supportive care for organ dysfunction, while attempts to block individual cytokines or signaling pathways have not consistently produced clinical benefit. Conventional nanocarriers can improve the stability, circulation and accumulation of drugs, but most release their cargo according to a predetermined schedule. That static approach may be poorly matched to a disease whose signals vary among patients, organs and disease stages. Stimulus-responsive platforms instead encode chemical or biological switches into a carrier, allowing oxidation, acidity, enzymes, microbial molecules or externally applied physical energy to alter its structure and release therapeutic cargo.

The most direct endogenous trigger is local acidification. Ischemia, impaired microcirculation, increased glycolysis and microbial metabolism can lower extracellular pH from the physiological range of 7.4 to approximately 6.5–6.8 at infected or injured sites. Hydrazone, acetal and ketal bonds can be hydrolyzed under mildly acidic conditions, while protonatable groups can reverse a carrier’s charge and weaken hydrophobic interactions. One preclinical system delivered a glucose transporter 1 inhibitor through a pH-responsive nanoparticle. Acid-triggered disassembly reduced macrophage glycolytic flux and shifted macrophages away from a pro-inflammatory phenotype, with reported improvements in organ injury and survival in septic animals. Such findings show the proposed mechanism in animals, not clinical efficacy.

Oxidative stress provides another controllable switch. Reactive oxygen species generated by activated neutrophils, macrophages and endothelial cells can reach micromolar concentrations in acute infection, compared with less than 100 nM in healthy tissue according to the review. Oxidation-sensitive thioketals, thioethers, selenium-containing bonds and boronate esters can therefore cleave or transform carriers in heavily oxidative environments. Some platforms have a dual function: their matrix scavenges reactive species while the carrier simultaneously releases an antioxidant, anti-inflammatory or immunomodulatory drug. A mouse model of sepsis-associated acute kidney injury, for example, used atorvastatin-loaded, mitochondria-targeted cerium oxide nanoparticles coated with an oxidation-sensitive polymer. The researchers reported reduced tubular-cell death, oxidative stress and inflammation, but no corresponding human trial was described.

Enzymes and microbial molecules can provide more infection-specific triggers. Matrix metalloproteinases, phospholipase A2, neutrophil proteases and bacterial lipases can destabilize peptide linkers or lipid matrices. A vancomycin-loaded solid lipid nanoparticle designed to bind bacterial lipase and hyaluronidase reduced the reported minimum inhibitory concentration against Staphylococcus aureus by twofold relative to free vancomycin and increased MRSA biofilm eradication fivefold in vitro. Other platforms recognize lipopolysaccharide, bacterial nucleic acids or pathogen enzymes. A phosphocholine-functionalized magnetic nanoadsorbent captured lipopolysaccharide in a septic rabbit model, while an inhalable DNase I and human serum albumin nanomotor cleared cell-free mitochondrial DNA from mouse lungs. These are preclinical results using model-specific signals, and the review cautions that accessibility, abundance and strain dependence vary among pathogens and infection sites.

External controls such as ultrasound offer precise release at a selected location but cannot readily cover the whole circulation of a systemic disease. In one sepsis-associated kidney injury model, a kidney-targeted liposome released the Toll-like receptor 4 inhibitor TAK-242 after low-intensity focused ultrasound, suppressing inflammatory and pyroptosis-related pathways in the animals. Another ultrasound-responsive formulation combined iron oxide particles and the compound FOT to support sonodynamic treatment in local infection and sepsis models. These systems are best suited, the review argues, to localized infection or protection of a particular organ. Light has limited penetration in deep tissue, while magnetic triggering faces difficulty in uniformly controlling particles throughout the body. Combining external control with endogenous sensing may therefore offer a more practical compromise.

The therapeutic cargo matters as much as the release mechanism. Responsive carriers have been used in preclinical studies to deliver antibiotics and antimicrobial peptides, anti-inflammatory and antioxidant drugs, small interfering RNA, microRNA and other immune-regulating agents. One infection-responsive nanoparticle co-loaded ciprofloxacin with an anti-inflammatory compound and used acidic pH and bacterial enzymes as release triggers; in a lethal bacterial challenge model, the researchers reported improved survival compared with controls. A separate hydrazone-linked prodrug combined the antimicrobial peptide omiganan with dexamethasone and was released through acidic conditions and hyaluronidase activity in infection and sepsis animals. Such designs can couple pathogen suppression with control of host inflammation, but they do not bypass the need for antibiotics or establish superiority in human sepsis.

Translation remains the central obstacle. Patient-derived cecal-ligation-and-puncture and endotoxin models reproduce important features of sepsis but do not fully reproduce the heterogeneous immune trajectories, multi-organ interactions and individual progression seen in patients. Nanoparticles must also avoid complement activation and rapid clearance, remain stable long enough to reach their targets, and then degrade into nontoxic, excretable products. Researchers need standardized production, pharmacokinetic and biodistribution studies, organ-specific clearance data and longer-term safety assessments. The review identifies manufacturing consistency as a particular problem for carriers combining multiple linkers, sensing elements and biological cargo, and it presents closed-loop systems combining real-time biosensing with artificial intelligence as a future concept rather than an established treatment.

The most plausible route described by the review is therefore not a single universal nanoparticle but a family of state-aware systems. A platform might sense acidity, oxidative stress and enzyme activity together, suppress inflammation early, and deliver immune-stimulating or gene-regulating cargo when signs of immunosuppression predominate. Boolean logic gates, adjustable linker density and sequential core-shell structures could reduce premature release and synchronize treatment with changing conditions. The central claim is that matching nanocarrier behavior to the biology of sepsis may improve the therapeutic window compared with static delivery. That benefit has been observed in selected animal and laboratory studies, while human validation, reliable endotype classification, closed-loop control, scalable manufacture and long-term safety remain unresolved.

Subject of Research: Review of stimulus-responsive nanoplatforms for targeted drug delivery and immune modulation in sepsis

Article Title: Pathological Signal‐Responsive Nanoplatforms for Sepsis: Integrating Biomarker Sensing With Spatiotemporal Drug Delivery and Immunomodulation

Article References: Liu, Y., Wang, K., Yan, K. K., Xu, Z., Zhao, X., Bai, X., Li, Z., & Wang, Y. (2026). Pathological Signal‐Responsive Nanoplatforms for Sepsis: Integrating Biomarker Sensing With Spatiotemporal Drug Delivery and Immunomodulation. Advanced Science, 13(58), Article e77718. https://doi.org/10.1002/advs.77718

Image Credits: AI Generated

DOI: 10.1002/advs.77718

Keywords: sepsis, stimulus-responsive nanoparticles, precision drug delivery, immunomodulation, nanomedicine, pH-responsive release, oxidative stress, microbial targeting, animal models, Pathological, Signal, Responsive

Sepsis was identified using Sepsis-3 criteria based on the Sequential Organ Failure Assessment score and quick SOFA criteria. The clinical analysis was descriptive rather than interventional, so the reported associations do not establish that any antibiotic regimen, patient characteristic, or dosing strategy caused an improvement or deterioration in outcomes.

Among the 196 positive cultures, the average hospital stay was 22 days. For patients who required intubation and transfer to the intensive care unit, the average stay rose to 41 days. Mortality was higher among patients aged 65 or older in both sexes. It was 8 of 37 men younger than 65, compared with 24 of 69 men aged 65 or older; among women, it was 4 of 44 younger patients and 8 of 46 aged 65 or older. Men had higher observed mortality in each age group, although the analysis did not demonstrate that age or sex caused the differences.

Cardiovascular disease was the most frequently recorded concomitant condition, reported in 38% of the study population, and diabetes mellitus was second at 24%. These findings describe the clinical profile of the analyzed group. As a single-center retrospective series based on blood cultures, the study did not include an untreated or non-sepsis comparison group and could not determine whether the observed lengths of stay, mortality, or resistance patterns were representative of other hospitals or populations.

The isolates included a broad range of bacteria, with Staphylococcus species predominating. The investigators also identified enterobacteria such as Klebsiella and Escherichia coli producing extended-spectrum beta-lactamases, along with Acinetobacter baumannii, Providencia stuartii, Citrobacter freundii, and Klebsiella oxytoca. The presence of ESBL-producing enterobacteria and A. baumannii complicated treatment choices in the study population, while the rare isolates demonstrated the etiological diversity among the sampled patients.

Resistance patterns differed among the organisms examined. Among the Gram-positive isolates, Staphylococcus haemolyticus showed 37% resistance to erythromycin and approximately 28% resistance to levofloxacin, while Enterococcus faecalis showed high resistance to ciprofloxacin and erythromycin. The authors associated the macrolide and fluoroquinolone resistance with the widespread use of these drug classes over preceding years. S. aureus had very low measured resistance in this sample, and no Gram-positive isolate was resistant to vancomycin, teicoplanin, or linezolid.

Among the Gram-negative organisms, ESBL-producing K. pneumoniae had the highest resistance to most tested penicillins and cephalosporins, and the pattern for A. baumannii was not substantially different. By comparison, K. oxytoca and Proteus mirabilis were susceptible to most antibiotics tested. These results identify resistance in the sampled blood-culture isolates; they do not determine which drugs were administered, whether prescribed therapy was appropriate for each isolate, or whether resistance was associated with an individual patient’s outcome.

The review also describes how sepsis-related physiological changes can alter antibiotic exposure. Capillary leak and fluid movement into interstitial spaces expand the distribution volume of hydrophilic drugs such as aminoglycosides, beta-lactams, glycopeptides, and lipopeptides, which can increase loading-dose requirements. Augmented renal clearance, commonly operationalized as creatinine clearance of at least 130 mL/min/1.73 m², may further accelerate elimination and has been associated with antibiotic underdosing. Hypoalbuminemia may increase the free fraction and clearance of highly protein-bound drugs such as ceftriaxone, potentially reducing drug exposure near the end of a dosing interval.

Within those pharmacological constraints, the authors discuss pharmacokinetic and pharmacodynamic dosing strategies. Prolonged or continuous beta-lactam infusions are intended to maintain concentrations above the pathogen’s minimum inhibitory concentration, while high loading doses and subsequent adjustments are described for vancomycin. High-dose, extended-interval regimens are presented for concentration-dependent drugs such as aminoglycosides and fluoroquinolones, with therapeutic drug monitoring particularly important for aminoglycosides. The study concludes that collaboration between physicians and clinical pharmacists is important for individualized antibiotic dosing, but the retrospective blood-culture data do not directly test these strategies or establish their clinical effectiveness in the included patients.

Subject of Research: Antibiotic resistance, pathogen distribution, and pharmacokinetic challenges in patients with sepsis

Article Title: Antibiotic treatment in sepsis–etiological structure, antibiotic resistance, and pharmacological challenges: A retrospective analysis

Article References: Kobakova, Y. A., Kirkova-Bogdanova, A. G., Vakrilova Becheva, M. S., Semerdzhieva, N. E., Ivanova, S. A., Chaneva, M. S., Atanasova, V. P., Bashov, I. E., Yonchev, Y. T., Todorov, R. F., Bashev, N. Z., Teneva, D. N., Danov, K. V., Moneva-Sakelarieva, M. G., Atliev, K. T., Kyosev, V. S., Pavlova, S. M., & Atanasov, P. Y. (2026). Antibiotic treatment in sepsis–etiological structure, antibiotic resistance, and pharmacological challenges: A retrospective analysis. Pharmacia, 73, Article e212551. https://doi.org/10.3897/pharmacia.73.e212551

Image Credits: AI Generated

DOI: 10.3897/pharmacia.73.e212551

Keywords: sepsis, blood cultures, antibiotic resistance, pathogen distribution, pharmacokinetics, pharmacodynamics, therapeutic drug monitoring, antimicrobial stewardship, retrospective study, Antibiotic, treatment, etiological

Patients had to remain in the intensive care unit for at least 24 hours. The investigators analysed each marker using the first value obtained after ICU admission, recorded at a median of 0.6 hours with an interquartile range of 0.5 to 4.5 hours, and, in parallel, the worst value within the first 24 hours. They compared discrimination using area under the receiver-operating-characteristic curve (AUC) and the DeLong test, overall and within severity strata and by admission source. The septic-shock approximation was based on vasopressor use and a first post-ICU venous lactate above 2 mmol/L.

With first measurements, CRP had an AUC of 0.504 (95% confidence interval, 0.439 to 0.567) and PCT had an AUC of 0.549 (0.486 to 0.614), both at or near chance-level discrimination. The NLR point estimate was below chance at 0.438 (0.371 to 0.501), although its confidence interval included 0.50. IL-6 had an AUC of 0.694 (0.631 to 0.756), and venous lactate had an AUC of 0.682 (0.622 to 0.742). Results were similar when the worst values within 24 hours were used, with all paired DeLong p values of at least 0.19.

Adding first IL-6 and lactate values to a partially window-aligned clinical model increased AUC from 0.782 to 0.808. The likelihood-ratio p value was 0.0004, while the DeLong comparison yielded p = 0.057. Performance was weaker among the 92 patients who did not meet the septic-shock approximation: AUC was 0.578 for IL-6 and 0.502 for lactate. The authors said this attenuation meant that restriction to a narrow, high-severity case mix could not be excluded as an explanation. By contrast, assay censoring appeared insufficient to explain the limited prognostic information associated with CRP and neutrophil-based indices: CRP and NLR were not significantly correlated with reconstructed SOFA, and only 1.3% of first CRP values reached the assay ceiling. These findings come from a retrospective analysis summarized in the abstract; they do not establish prospective performance beyond the included high-acuity ICU population, and the authors called for prospective evaluation.

Subject of Research: Retrospective comparison of inflammatory markers, IL-6 and lactate for predicting 28-day mortality in high-acuity ICU sepsis

Article Title: Head-to-head comparison of routinely ordered inflammatory markers, interleukin-6 and lactate for 28-day mortality in high-acuity ICU sepsis: a predominantly vasopressor-treated cohort

Article References: Feng, L., Xie, Q., Wang, Y., & Li, K. (2026). Head-to-head comparison of routinely ordered inflammatory markers, interleukin-6 and lactate for 28-day mortality in high-acuity ICU sepsis: a predominantly vasopressor-treated cohort. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14472-3

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14472-3

Keywords: sepsis, intensive care, 28-day mortality, interleukin-6, venous lactate, C-reactive protein, procalcitonin, neutrophil-to-lymphocyte ratio, vasopressors, retrospective cohort, Head-to-head, comparison

Active rewarming was implemented at the discretion of treating clinicians rather than assigned as part of a standardized protocol. The analysis found no association between active rewarming and 90-day mortality, the incidence of acute respiratory distress syndrome or acute kidney injury. Because this was an observational study and rewarming was not randomly assigned, the results do not establish that active rewarming has no effect; they show that receipt of the intervention was not associated with improved outcomes in this cohort.

Group-based trajectory modeling identified four distinct temperature-trajectory phenotypes. Faster restoration of temperature, whether it occurred spontaneously or with active warming, tended to be associated with better survival. By contrast, persistent hypothermia and incomplete rewarming were linked to elevated 28-day mortality risk. The researchers also conducted subgroup and sensitivity analyses, but the abstract does not report the details of those analyses or their results. Overall, the temperature-recovery trajectories appeared more closely associated with prognosis than whether patients received active rewarming.

The findings are bounded by the study’s retrospective, multicenter design and by clinician-directed selection of active rewarming. The supplied abstract also does not describe the distribution of patients among the four trajectory phenotypes, potential confounding factors, or uncertainty for the reported associations. It presents no clinical trial number because the work was not a clinical trial. The authors concluded that prospective studies using standardized rewarming protocols are needed to refine temperature-management strategies for patients with hypothermic sepsis.

Subject of Research: Associations of active rewarming and temperature-recovery trajectories with mortality in hypothermic sepsis

Article Title: Temperature trajectories, active rewarming, and mortality in hypothermic sepsis: a multicenter retrospective cohort study

Article References: Zhang, X., Chen, L., Xu, Y., Chen, F., Sun, R., Xu, Y., Xu, Y., Xu, J., & Chen, H. (2026). Temperature trajectories, active rewarming, and mortality in hypothermic sepsis: a multicenter retrospective cohort study. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14505-x

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14505-x

Keywords: hypothermic sepsis, active rewarming, temperature trajectories, 28-day mortality, sepsis, retrospective cohort, intensive care, acute respiratory distress syndrome, acute kidney injury, Temperature, trajectories, active

Sepsis continues to carry a high mortality burden despite advances in antimicrobial therapy, source control, hemodynamic resuscitation, and organ support. Its biological course involves not only excessive inflammation but also changes in immune regulation, oxidative balance, mitochondrial function, endothelial integrity, metabolism, coagulation, and organ homeostasis. Innate immune receptor activation can engage NF-κB, MAPK, inflammasome, and related pathways, leading to mediators such as TNF-α, IL-1β, IL-6, and HMGB1. Yet these signals operate within an interconnected network, making it difficult to expect consistent benefit from suppressing one cytokine alone.

Mitochondrial dysfunction is one of the central links in that network. Impaired oxidative phosphorylation, excessive reactive oxygen species, ATP depletion, and mitochondrial damage can intensify inflammatory signaling and cellular injury, while the resulting tissue stress can further disrupt mitochondrial function. The editorial therefore treats mitochondrial quality control—including biogenesis, dynamics, and mitophagy—as a possible therapeutic target rather than a secondary feature of sepsis. Experimental work on Xiangdan injection reported attenuation of inflammatory responses and cardiac injury together with mitochondrial protection in an experimental sepsis model. However, many preclinical studies define mitochondrial protection through morphology or selected protein markers rather than direct measurements of respiration, ATP production, membrane potential, mitophagy flux, or biogenesis. The decisive question is therefore not only whether a treatment lowers oxidative-stress markers, but whether it restores cellular bioenergetic function.

Another strand of the collection focuses on the vascular consequences of immune activation. Neutrophil extracellular traps, or NETs, can contribute to endothelial damage, coagulation activation, increased vascular permeability, immunothrombosis, and impaired microcirculation. Because organ injury depends on perfusion, oxygen delivery, and microvascular blood flow as well as cytokine concentrations, therapies that modify interactions between inflammatory cells and the vasculature may complement direct cytokine inhibition. In experimental sepsis, docosahexaenoic acid was associated with reduced inflammatory responses, neutrophil adhesion, and microcirculatory abnormalities. The same study reported increased formation of pro-resolving mediators, but that measurement used enzyme immunoassays with limited selectivity in complex biological samples; the authors therefore say the result requires confirmation by LC-MS/MS. They also argue that future NET-directed studies should test whether such treatments improve organ perfusion and clinical outcomes without weakening antimicrobial host defense.

The editorial also highlights a central problem with the cytokine-storm framework: hyperinflammation and immunosuppression can coexist, and a patient may move from one state to another during the course of illness. A systematic review and network meta-analysis in the collection synthesized 76 randomized trials involving 22,194 patients and assessed several immunomodulatory interventions. Ulinastatin, ulinastatin combined with thymosin-α1, and polyunsaturated fatty acids showed potential benefits, but heterogeneity and limitations in the certainty of the evidence temper those results. The relevant clinical question may thus be not whether inflammation should always be suppressed, but which biological phenotype requires treatment and when. Patients dominated by hyperinflammation may be biologically distinct from those with impaired immune competence, and treating both as a single syndrome could produce inconsistent results.

Metabolic and cardiovascular pharmacology broadens the potential toolkit beyond conventional anti-inflammatory agents. In a randomized controlled study of patients with septic shock, anisodamine hydrobromide was associated with lower 28-day mortality, with lactate-related variables contributing to the observed effect. The result is consistent with the possibility that improving microcirculation and tissue perfusion may be therapeutically important, although the editorial does not present the study as definitive evidence for routine use. A study of the GLP-1 receptor agonist liraglutide during sepsis reported effects on retinal endothelial function and oxidative stress, indicating that metabolic signaling may influence vascular injury during systemic inflammation. In a separate real-world cohort of patients with diabetic kidney disease, finerenone was associated with lower risks of sepsis-induced cardiomyopathy and mortality than steroidal mineralocorticoid receptor antagonists. Because that finding was observational, it cannot establish causality, but it supports prospective evaluation of drugs whose host-directed effects may extend beyond their original indications.

Natural products and drug repurposing form a substantial part of the research collection, which included work on artesunate, Dachengqi Decoction, Xiangdan injection, sophoridine, ganoderic acid A, Handelin, and docosahexaenoic acid. Their potential advantage is multitarget activity across inflammatory, oxidative, mitochondrial, metabolic, and cell-death pathways, all of which contribute to sepsis. The same complexity creates translational obstacles: variable composition, uncertain bioavailability, poorly characterized pharmacokinetics, difficulty achieving clinically relevant concentrations, and limited randomized evidence. Some work is beginning to move beyond pathway association. Multi-omics combined with experimental validation identified TNF-α as a candidate target in sepsis-related liver injury in a ganoderic acid A study, while proteomic profiling of Handelin nominated the proteasome subunits PSMA7 and PSMB2 as possible targets in the regulation of proteostasis. A separate study integrated metabolomics, drug-metabolism and pharmacokinetic analysis, network pharmacology, molecular modeling, cellular experiments, and animal work to investigate artesunate-induced nephrotoxicity. That model does not directly demonstrate sepsis efficacy, an important distinction the editorial emphasizes.

Evidence for artesunate itself was more directly related to sepsis in a systematic review and meta-analysis of animal studies, which reported beneficial effects on survival and inflammatory or organ-injury outcomes. The authors nevertheless note heterogeneity and limitations in the preclinical evidence. The broader lesson is that candidate interventions need to progress from associations in omics or pathway analyses to validated molecular targets, measurable exposure, target engagement, functional efficacy, and eventual clinical testing. This applies particularly to combination therapy. Combining agents merely because they act through different mechanisms is less persuasive than selecting combinations that correct complementary abnormalities in patients whose biology supports that strategy.

Organ-specific toxicity and precision dosing add another dimension to sepsis pharmacology. Among patients with sepsis, higher vancomycin concentrations were associated with increased risk of acute kidney injury, and hyperlactatemia modified that relationship. The finding indicates that the same drug exposure may have different effects depending on a patient’s physiological state. A systematic review comparing vasopressin with norepinephrine found potential renal advantages for vasopressin, including lower serum creatinine and reduced renal-replacement therapy, although effects on acute kidney injury were less consistent. Such findings reinforce that advances in sepsis treatment require not only new drugs but also better optimization of existing medicines according to organ function, metabolic condition, infection characteristics, and measured exposure.

Several gaps remain before this precision approach can be applied reliably. Trials have not yet incorporated biological heterogeneity enough to identify which patients are most likely to benefit from particular interventions. Biomarkers of inflammation, immune competence, endothelial injury, mitochondrial dysfunction, and metabolism may eventually help classify patients, while the timing of treatment is likely to matter because an intervention that helps during early hyperinflammation may be ineffective or harmful during later immunosuppression. Mechanistic claims also require stronger validation than molecular docking, transcriptomic associations, or pathway enrichment; target engagement, genetic manipulation, rescue experiments, and functional assays provide more direct causal evidence. Preclinical models should better represent differences in age, sex, diabetes, obesity, chronic kidney disease, immunosuppression, infection characteristics, antibiotics, fluid resuscitation, and vasopressor use. The editorial’s conclusion is not that a universal anti-inflammatory drug is unnecessary, but that broader suppression should give way to interventions selected for the right biological phenotype at the right time. By integrating immune phenotyping with mitochondrial, vascular, metabolic, pharmacokinetic, and organ-specific information, sepsis therapy may become better able to interrupt self-amplifying injury while preserving antimicrobial defense.

Subject of Research: Precision pharmacological treatment of sepsis across inflammatory, mitochondrial, vascular, and metabolic pathways

Article Title: Editorial: Pharmacological modulation of inflammatory pathways in sepsis: targeting the cytokine storm and septic oxidative stress

Article References: Zhang, G., VanBramer, H., Wang, D., Prasad, A., & Meng, Q. (2026). Editorial: Pharmacological modulation of inflammatory pathways in sepsis: targeting the cytokine storm and septic oxidative stress. Frontiers in Pharmacology, 17, Article 2021242. https://doi.org/10.3389/fphar.2026.2021242

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.2021242

Keywords: sepsis, cytokine storm, oxidative stress, mitochondrial dysfunction, immunomodulation, microcirculation, natural products, precision medicine, drug repurposing, Editorial, Pharmacological, modulation

Researchers at the Affiliated Hospital of Nantong University reviewed consecutive adults with Sepsis-3-defined sepsis who received clinician-directed mNGS between January 2022 and November 2025. Of 251 patients screened, 142 were excluded because essential clinical information was unavailable, leaving 109 patients for the overall analysis. Fifty-eight were classified as clinically actionable, meaning that antimicrobial treatment was documented as being modified in response to the mNGS findings, while 51 were classified as non-actionable. A further 37 patients lacked the paired post-report C-reactive protein measurement required for the complete variable-selection procedure, so the final model-development cohort contained 72 patients, including 39 actionable and 33 non-actionable cases.

The mNGS tests detected at least one microorganism in 98 of the 109 patients, compared with 50 detections by conventional culture. The detection yield was 94.8% in the actionable group and 84.3% in the non-actionable group, whereas culture detected organisms in 53.4% and 37.3% of those respective groups. The sequencing assay also produced a descriptively broader organism spectrum, including 12 fungal taxa and 18 DNA or RNA viral taxa not recovered by culture. These comparisons were made at the patient level rather than through uniformly matched specimens, and the clinical reference assessment incorporated mNGS results, so the estimated concordance does not demonstrate that mNGS was independently superior for diagnosis.

Clinical actionability was defined through a prespecified process by two experienced intensive-care physicians who had not cared for the included patients. They reviewed the sampling site, clinical syndrome, imaging, conventional microbiology, treatment history, biological plausibility and the contemporaneous rationale in medical records; disagreements were resolved by consensus. Detection alone did not qualify a result as actionable. Among the 58 actionable cases, 39 involved antimicrobial escalation, eight de-escalation, eight antifungal modification, two antiviral modification and one discontinuation. Among the 51 non-actionable cases, therapy remained unchanged in 49, while two patients underwent modifications not documented as attributable to mNGS. The endpoint therefore measured a pragmatic treatment change rather than whether the change was appropriate or beneficial.

Six prespecified changes measured between the pre-mNGS assessment and the first eligible reassessment after report availability were evaluated: white blood cell count, neutrophil percentage, log-transformed procalcitonin, log-transformed C-reactive protein, Sequential Organ Failure Assessment score and Acute Physiology and Chronic Health Evaluation II score. These follow-up measurements were obtained before any antimicrobial modification classified as the outcome, although clinicians could see the individual laboratory results while making treatment decisions. The timing was not standardized, and the interval between the report and repeat testing could not be reliably reconstructed for all patients. Among the six candidates, only the change in white blood cell count, designated dWBC, passed the univariable screening threshold and was retained by the subsequent least absolute shrinkage and selection operator procedure.

The baseline model used information available before mNGS and included age, diabetes, baseline white blood cell count and baseline log-transformed procalcitonin. The dynamic model added dWBC and was intended for use only after the sequencing report became available. In the 72-patient development cohort, the apparent area under the receiver operating characteristic curve rose from 0.744 for the baseline model to 0.816 after adding dWBC, while the apparent Brier score fell from 0.199 to 0.173. However, a paired DeLong comparison did not reach statistical significance, with P = 0.088, so the numerical improvement did not establish that the dynamic model was superior. In the broader cohort, dWBC remained positively associated with actionability in univariable analysis, but it was no longer statistically significant after adjustment for the four baseline variables in a post hoc refitting, with an adjusted odds ratio of 1.06, 95% confidence interval 0.99–1.13 and P = 0.092.

Internal assessment used 1,000 nonparametric bootstrap resamples, with all resamples successfully fitted. Because the analysis refitted the finalized predictor sets without repeating univariable screening and variable selection, the correction was conditional on those choices and could underestimate total development optimism. The optimism-corrected AUCs were 0.697 for the baseline model and 0.768 for the dynamic model, while the corrected Brier scores were 0.228 and 0.207, respectively. Corrected calibration slopes were 0.783 and 0.732, indicating some residual overfitting. An exploratory decision-curve analysis based on uncorrected predictions suggested greater net benefit for the dynamic model at parts of the examined threshold range, but it was not optimism-corrected and did not establish clinical utility or patient benefit.

The investigators also tested whether the relationship between dWBC and actionability differed according to baseline white blood cell count. Among patients with a baseline count of 12 × 10⁹/L or less, the adjusted odds ratio for dWBC was 1.15, with a 95% confidence interval of 0.95–1.38 and P = 0.144. Above that threshold, the adjusted odds ratio was 1.16, with a 95% confidence interval of 1.01–1.34 and P = 0.034. Although statistical significance appeared only in the higher-count stratum, the point estimates were nearly identical and their confidence intervals overlapped. A formal continuous interaction test was not significant, with P = 0.494, providing no evidence that the association varied according to baseline white blood cell count. The subgroup analysis was exploratory because the strata were small.

Several limitations constrain interpretation. The study was small, retrospective and confined to one center, and the exclusion of 142 screened patients and the further restriction to 72 complete cases may have introduced selection bias. Differences in sex and specimen distribution also limit generalizability, while requiring an eligible post-report measurement before treatment modification may have created measurement-availability or survivor bias. Respiratory specimens were more common in the actionable group, making anatomical sampling site a potential confounder that the small model could not assess reliably. Treatment changes were attributed retrospectively, and incomplete documentation, hindsight bias, shared institutional practice and the absence of a counterfactual prevented causal conclusions. The authors also did not assess treatment appropriateness, broader effects on infection control or public health, cost-effectiveness, or patient-centered outcomes. Prospective multicenter validation with standardized repeat measurements and a prespecified post-report pathway is therefore needed before the model could support clinical decisions.

Subject of Research: Development of a post-report sepsis model linking peri-report white blood cell changes to mNGS-associated antimicrobial modification.

Article Title: Predicting clinically actionable metagenomic next-generation sequencing in sepsis through dynamic inflammatory trajectories

Article References: Ding, X., Bao, P., Huang, X., Xu, R., Tang, X., Wu, Y., Cui, X., & Lu, Y. (2026). Predicting clinically actionable metagenomic next-generation sequencing in sepsis through dynamic inflammatory trajectories. Frontiers in Microbiology, 17, Article 1949602. https://doi.org/10.3389/fmicb.2026.1949602

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1949602

Keywords: sepsis, metagenomic sequencing, white blood cell count, clinical prediction model, antimicrobial stewardship, infectious diseases, diagnostic actionability, inflammatory trajectories, intensive care, Predicting, clinically, actionable

Sepsis-associated lung injury arises from a dysregulated response to infection and frequently involves acute respiratory distress syndrome, or ARDS. A meta-analysis cited in the review estimated that ARDS develops in 34.0% of patients with sepsis, with a 95% confidence interval of 29.0% to 39.3%. Lysosomes and related endolysosomal compartments sit directly at the host–pathogen interface: acidic phagolysosomes support microbial killing in macrophages, lamellar bodies participate in surfactant biology in type II alveolar epithelial cells, and lysosomal degradation and trafficking contribute to endothelial repair and fibroblast matrix turnover. Infection-related stress could therefore compromise both pathogen control and the cellular processes needed to preserve the lung’s gas-exchange barrier. The proposed link between these defects and lysosomal pH remains incompletely mapped, however, particularly in human disease.

Central to the hypothesis is the V-ATPase complex, which hydrolyzes ATP to pump protons into lysosomal compartments and create the acidic conditions required by many lysosomal hydrolases. V-ATPase is also a signaling platform influenced by nutrient sensing, trafficking, complex assembly and stress. In macrophages, the ATP6V0D2 subunit has been associated with autophagosome–lysosome fusion and restrained inflammasome activation during bacterial challenge, illustrating that its role cannot be reduced to setting basal acidity. The mTORC1–TFEB pathway provides another connection between nutrient status, lysosomal biogenesis, autophagy and stress adaptation, but the review cautions that TFEB does not uniformly increase every acidification-related transporter across cell types or conditions. Similarly, PI3K–AKT signaling may influence lysosomal resilience indirectly through metabolism, mitochondrial function and membrane stability; a direct AKT-driven alkalinization mechanism has not been established in septic lung cells.

Acidification also depends on the movement of counterions. ClC-7 is a late-endosomal and lysosomal chloride–proton exchanger that requires the accessory protein OSTM1 for stability. Recent work indicates that ClC-7 promotes chloride accumulation in mature phagosomes, hydrolase activity and phagosome resolution. Its influence on bulk lysosomal acidity is more complicated, with genetic studies reporting altered chloride distribution and storage phenotypes without consistent loss of acidification. The authors therefore describe ClC-7 primarily as a regulator of luminal chloride with context-dependent effects on pH. TMEM175, an endolysosomal potassium channel implicated in membrane potential and pH stability, presents another unresolved question. Models disagree over whether it conducts protons or predominantly potassium, and the evidence comes largely from heterologous systems or non-septic neurodegeneration-related models. Its role in pulmonary sepsis has not been directly tested.

A further complication is that lysosomal pH is not equivalent to every other sign of lysosomal dysfunction. The review distinguishes increased luminal pH from membrane permeabilization, reduced hydrolase activity, impaired autophagic flux and changes in lysosomal mass or biogenesis. Markers such as LC3-II, p62, cathepsin abundance and LysoTracker intensity can each be altered by organelle number, trafficking, cargo load, membrane damage or probe behavior. LC3-II accumulation, for example, may reflect either increased autophagy or failure of lysosomal degradation. Direct acidity evidence ideally comes from compartment-targeted ratiometric probes calibrated in situ with ionophores and defined-pH buffers. Single-wavelength acidotropic dyes are useful for locating lysosomes but cannot, by themselves, provide stand-alone pH values. This distinction is especially important in sepsis, where cell death, systemic acid–base disturbances and changes in organelle abundance can confound commonly used readouts.

Microbial and host signals could disturb lysosomal homeostasis in several ways. Lipopolysaccharide activates TLR4–MyD88/TRIF and NF-κB inflammatory signaling in pulmonary immune and structural cells, but available evidence does not show universal repression of acidification machinery. Instead, changes in trafficking, cytokines, nitroxidative stress, metabolism and autophagic flux may indirectly alter lysosomal function. Tumor necrosis factor alpha, interleukin-1 beta and interleukin-6 can promote reactive oxygen species, mitochondrial injury, metabolic reprogramming and dysregulated autophagy–lysosome flux. Persistent stress may reduce ATP-dependent proton pumping, disturb membrane stability or overload degradative capacity. Lysosomal membrane permeabilization can then release cathepsins and other contents into the cytosol. Cathepsin B has been linked to NLRP3 inflammasome activation and pyroptosis-related signaling in experimental sepsis-associated lung injury, but its contribution depends on the cell and stimulus. Reduced lysosomal acidity is thus best considered a possible permissive amplifier working with mitochondrial reactive oxygen species, potassium efflux, danger signals and defective autophagic clearance, not as the sole trigger of inflammation.

The consequences may differ sharply among the cells that form and defend the alveolar–capillary barrier. In macrophages and dendritic cells, impaired lysosomal or phagolysosomal function may reduce microbial killing, cargo degradation and antigen processing, which is why indiscriminate suppression of lysosomal activity could be counterproductive. In epithelial and endothelial cells, lysosomal injury may contribute to protease release, junctional disruption, vascular leakage and defective repair through convergent mechanisms, but direct pH-to-junction causality has not been established. Type II epithelial lamellar bodies have specialized functions in calcium handling and surfactant secretion that should not be equated automatically with canonical lysosomes. During persistent or non-resolving injury, impaired lysosomal degradation may also hinder removal of damaged mitochondria, oxidized lipids and mitochondrial DNA, potentially sustaining TLR9, cGAS–STING and NLRP3 pathways. Evidence linking lysosomal pH to fibrotic matrix remodeling comes mainly from pulmonary fibrosis research, and direct human evidence in post-sepsis lung injury remains limited.

Mitophagy and inter-organelle proton transfer add another possible dimension. Mitochondria and lysosomes communicate through mitophagy, reactive oxygen species, lipid transfer and direct membrane contact sites. Mitochondrial injury can restrict the ATP and redox conditions needed for V-ATPase activity, while defective lysosomal trafficking can reduce mitochondrial clearance and allow dysfunctional mitochondria to accumulate. Studies in yeast and human cells have now shown proton transfer from mitochondria to lysosomes at contact sites, with experimental organelle tethering improving acidification, cargo degradation and autophagic flux. These observations are mechanistically relevant but were not conducted in septic lung models, and no cited sepsis study simultaneously quantified mitochondria–lysosome contacts and lysosomal pH. The relative contributions of V-ATPase and counterion dysfunction, membrane damage and contact-mediated proton transfer therefore remain unresolved in this disease context.

Because the model is not yet ready for clinical application, the authors propose measured rather than diagnostic biomarker panels. A bronchoalveolar lavage fluid strategy could combine calibrated lysosomal pH in recovered alveolar macrophages, intralysosomal cathepsin B and D activity, cytosolic cathepsin redistribution, and extracellular cathepsin or LAMP1/2. A high-pH, low-hydrolase phenotype with limited extracellular release would be consistent with impaired acidification, whereas cathepsin redistribution and extracellular release would favor membrane injury. A less invasive blood panel could assess calibrated pH in circulating monocytes, cathepsin localization and autophagic flux, although neither LC3-II nor p62 alone would establish alkalinization. These are proposed test strategies rather than validated signatures and would require longitudinal cohorts, comparison with established measures such as oxygenation, lactate, procalcitonin and inflammatory markers, and analysis of whether blood and lung samples identify the same phenotypes.

Therapeutic development would likewise need to distinguish disease phase and molecular phenotype. An alkalinization-dominant state with impaired phagocyte degradation could justify testing controlled restoration of acidification or lysosomal biogenesis. One compound, ZLN005, increased lysosomal acidity and biogenesis in phagocytes and improved bacterial killing and survival in experimental polymicrobial sepsis, but its effects also involved PGC-1α, PI3K and TFEB, preventing attribution of benefit solely to pH correction. Selectively applicable clinical V-ATPase activators are not established. Conversely, a membrane-injury phenotype without clear alkalinization would support testing membrane stabilization or carefully titrated protease modulation, since broad cathepsin inhibition could interfere with microbial killing and tissue remodeling. Future experiments should combine spatial or single-cell identification of lysosomal stress with calibrated pH measurement, hydrolase assays, membrane-permeabilization tests and flux analysis in infection models, co-cultures, organoids and precision-cut lung slices. The defensible conclusion is that lysosomal pH homeostasis deserves testing as a mechanistic and stratification axis for sepsis-associated lung injury, not that it has already been established as a precision-medicine target.

Subject of Research: Lysosomal pH dysregulation as a candidate amplifier of sepsis-associated lung injury

Article Title: Lysosomal pH dysregulation as a candidate signal-amplifying hub in sepsis-associated lung injury

Article References: Tu, G., Deng, Z., Li, Y., Tian, P., Shen, Y., Deng, Y., Zhang, Z., Liu, Y., Zhong, S., Wu, J., Zhu, H., Ma, J., & Li, J. (2026). Lysosomal pH dysregulation as a candidate signal-amplifying hub in sepsis-associated lung injury. Frontiers in Cellular and Infection Microbiology, 16, Article 1953667. https://doi.org/10.3389/fcimb.2026.1953667

Image Credits: AI Generated

DOI: 10.3389/fcimb.2026.1953667

Keywords: lysosomal pH, sepsis, lung injury, ARDS, V-ATPase, inflammation, autophagy, mitochondria–lysosome crosstalk, host–pathogen interaction, Lysosomal, dysregulation, candidate

The cohort included children with impaired immune systems and serious infection requiring intensive care. Malignancy was the most common underlying condition, accounting for 56.5% of cases, and 76.1% of the patients had septic shock. Sepsis and septic shock were classified using the 2024 Phoenix Sepsis Score criteria.

The researchers used multivariate analysis to identify factors associated with mortality and receiver operating characteristic curve analysis to assess the score’s ability to distinguish survivors from nonsurvivors. For every one-point increase in the maximum score during the first 24 hours, the adjusted odds of death increased substantially. The study reported an adjusted odds ratio of 8.20 per point, with a 95% confidence interval of 1.35–49.69 and a p-value of 0.022.

This was the strongest independent prognostic association identified in the analysis. The confidence interval was wide, indicating substantial uncertainty around the estimated size of the association, an important limitation given the retrospective design and cohort size of 92 patients. The investigators also explored a Phoenix Sepsis Score threshold of 8 or higher.

These values describe performance in the study population only. The researchers characterized the threshold as exploratory, and their results do not establish that a score of 8 can be applied as a universal clinical rule or that the score itself causes changes in outcomes. The findings support further evaluation of the Phoenix Sepsis Score as a prognostic marker in immunocompromised children with sepsis and septic shock, but external validation is required.

Because the study was retrospective and confined to one pediatric intensive care unit in Thailand, the results do not establish performance in other hospitals, patient populations or healthcare settings. The clearest supported conclusion is that higher maximum Phoenix Sepsis Scores were strongly associated with mortality in this high-risk cohort, while the proposed cutoff of 8 remains an exploratory finding awaiting validation in independent populations.

Subject of Research: Prognostic performance of the Phoenix Sepsis Score in immunocompromised children with sepsis and septic shock

Article Title: Phoenix Sepsis Score as a prognostic tool for mortality in immunocompromised children with sepsis and septic shock: a six-year study from a Thai referral center

Article References: Menorath, P., Niamsanit, S., Nanthavongsa, S., Uppala, R., Sitthikarnkha, P., & Sriruengrat, D. (2026). Phoenix Sepsis Score as a prognostic tool for mortality in immunocompromised children with sepsis and septic shock: a six-year study from a Thai referral center. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14456-3

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14456-3

Keywords: Phoenix Sepsis Score, pediatric sepsis, septic shock, immunocompromised children, mortality prediction, prognosis, pediatric intensive care, Thailand, retrospective cohort, risk stratification, Phoenix, Sepsis

The research team, led by Youjia Hao and colleagues in the Department of Critical Care Medicine at Beijing Jishuitan Hospital, Capital Medical University, systematically searched PubMed, EMBASE, Web of Science, and the Cochrane Library from inception to January 2025. They included prospective studies reporting resistin levels in relation to mortality or disease severity in sepsis. After rigorous screening, eighteen studies encompassing 2,197 sepsis patients made it into the final analysis. The review was registered with PROSPERO under CRD42025639277, ensuring methodological transparency from the outset.

Resistin itself has an intriguing biology. Originally named for its proposed role in insulin resistance, the molecule is now recognized as a potent driver of inflammation. Unlike adipose-derived resistin in rodents, human resistin circulates largely from immune cells, and it amplifies the release of tumor necrosis factor and interleukin-6 while promoting endothelial dysfunction. In the context of sepsis, where a dysregulated host response to infection cascades into organ failure, resistin’s rise seems to mirror the intensity of that inflammatory storm.

The meta-analysis quantified this relationship using standardized mean differences, a statistical measure that allows comparison across studies employing different laboratory assays. Resistin levels were significantly higher in non-survivors than in survivors, with a pooled standardized mean difference of 0.83 and a 95 percent confidence interval spanning 0.31 to 1.36. Patients with septic shock, the most severe form of sepsis characterized by profound circulatory collapse, showed even more striking elevations compared to those without shock, with a standardized mean difference of 1.34 and a confidence interval of 0.62 to 2.05. Compared with healthy controls, sepsis patients overall displayed substantially elevated resistin, yielding a standardized mean difference of 1.99 with a confidence interval of 1.63 to 2.35.

When the researchers arranged these findings along the spectrum of disease, a stepwise gradient emerged. Healthy controls occupied the lowest rung, followed by patients with non-shock sepsis, then those with septic shock, and finally non-survivors at the top. This ordered pattern suggests that resistin is not merely a binary flag of infection but a quantitative readout of inflammatory burden, potentially offering clinicians a way to gauge how far along the severity curve a patient sits at the moment of admission.

The team also assessed resistin’s diagnostic accuracy using areas under the receiver operating characteristic curve, a metric that ranges from 0.5, equivalent to random guessing, to 1.0, representing perfect discrimination. The pooled AUC for predicting mortality was 0.68 with a confidence interval of 0.64 to 0.73. In adult patients specifically, the pooled AUC rose to 0.70 with a confidence interval of 0.65 to 0.75. These values indicate moderate discriminatory power, meaning resistin alone cannot reliably sort survivors from non-survivors but may contribute meaningful prognostic information when combined with clinical scoring systems.

However, the analysis was not without caveats. Substantial heterogeneity pervaded the main pooled analyses, reaching an I-squared value of 89 percent, a statistical signal that the included studies varied considerably in their populations, assays, and timing of measurement. The 95 percent prediction intervals for sepsis-related mortality, 28-day mortality, and septic shock all included zero, indicating that the direction of the effect in future studies remains uncertain. Subgroup and meta-regression analyses were conducted to explore potential effect modifiers, but the authors acknowledge that these exploratory tools cannot fully resolve the underlying variability.

The authors conclude that admission resistin levels may be associated with disease severity and adverse outcomes in sepsis, but they are careful to emphasize that clinical utility remains unproven. Their recommendation is unambiguous: large, multicenter prospective studies are needed before resistin can be integrated into routine prognostic assessment. In the interim, the molecule joins a growing list of inflammatory biomarkers, including procalcitonin and interleukin-6, whose prognostic value in sepsis is promising but not yet definitive.

For a condition that kills swiftly and often silently, any tool that sharpens early risk stratification carries enormous clinical weight. Resistin’s stepwise rise across the sepsis spectrum offers a compelling biological narrative, one that aligns with the broader understanding of sepsis as a crescendo of immune dysregulation. Whether this cytokine will earn a place beside lactate and procalcitonin in the emergency department or intensive care unit remains to be seen, but this meta-analysis provides the strongest synthesis to date of its prognostic signal.

The study, published open access under a Creative Commons license, received no specific funding and involved no competing interests. As sepsis continues to challenge clinicians worldwide, biomarkers like resistin represent a frontier where laboratory science meets bedside decision-making, and the coming years of multicenter validation will determine whether this inflammatory messenger becomes a standard component of the sepsis prognostic toolkit.

Subject of Research: The prognostic value of resistin in patients with sepsis: a systematic review and meta-analysis

Article Title: The prognostic value of resistin in patients with sepsis: a systematic review and meta-analysis

Article References: Hao, Y., Bai, Y., Zhang, L., Cui, Z., Liu, C., & Wang, H. (2026). The prognostic value of resistin in patients with sepsis: a systematic review and meta-analysis. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14293-4

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14293-4

Keywords: resistin, sepsis, biomarker, meta-analysis, septic shock, mortality, inflammation, BMC Infectious Diseases, prognostic, value, patients, systematic