Serum HMGB1 as a biomarker and predictive model for pediatric septic shock: a cohort study

Introduction

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection (1). In children, sepsis remains a leading cause of critical illness and mortality worldwide. Septic shock represents the most severe manifestation, characterized by circulatory and cellular/metabolic dysfunction severe enough to substantially increase mortality. While the term “severe sepsis” was previously used, current international guidelines (Sepsis-3, 2016) have discontinued this terminology, recognizing that all sepsis inherently implies organ dysfunction (2).

The pathogenesis of sepsis involves a complex dysregulated immune response to infection (3). Pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) bind to pattern recognition receptors (PRRs), triggering inflammatory mediator cascades (4). This excessive inflammatory response can lead to widespread tissue damage and organ dysfunction. High mobility group box 1 (HMGB1), a nuclear protein released during cellular stress and tissue injury, has emerged as a critical late mediator in this inflammatory cascade (5). Unlike early cytokines, HMGB1 is released hours after the initial insult and can perpetuate inflammation by activating multiple downstream pathways, including Toll-like receptors and the receptor for advanced glycation end products (RAGE) (6). This chain of events can escalate into life-threatening cytokine storms.

Studies have demonstrated that serum HMGB1 levels are significantly elevated in both experimental and clinical sepsis. In adult patients, elevated HMGB1 concentrations correlate with disease severity and adverse outcomes (7). Animal studies have shown that HMGB1 inhibition or neutralization can improve survival in septic mice, even when administered hours after the septic insult (8). This delayed kinetic profile of HMGB1 release—peaking 16–32 hours after initial infection—presents a potential therapeutic window and suggests its utility as a prognostic biomarker (9). However, despite these promising findings in adults and experimental models, the clinical value of HMGB1 as a prognostic biomarker in pediatric sepsis remains inadequately investigated (10).

This study aimed to address this knowledge gap by: (I) evaluating the prognostic value of serum HMGB1 levels for predicting septic shock in children with sepsis; (II) identifying independent risk factors for shock development; and (III) developing and internally validating a clinical prediction model combining HMGB1 with other biomarkers for early risk stratification. Given the limited evidence in pediatric populations and the potential clinical utility of early risk identification, this investigation sought to determine whether HMGB1 measurement could enhance clinical decision-making in pediatric sepsis management. We present this article in accordance with the TRIPOD reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-703/rc).

Methods

Case sources and grouping

Inclusion criteria

Subjects were enrolled if complete clinical records were available for review, were between 28 days and 18 years of age, and conformed to the diagnostic criteria outlined in the 2015 Chinese Expert Consensus on Pediatric Sepsis Shock Diagnosis and Treatment (11). During the study period [2022–2023], this consensus represented the standard of care at our institution, although we acknowledge that international guidelines have since evolved (Sepsis-3, 2016; Phoenix criteria, 2024).

Exclusion criteria

Subjects were excluded if they were under 28 days of age, had admission durations of less than 12 hours, had incomplete clinical data, or if their family members declined to sign informed consent forms. We also excluded patients with pre-existing conditions known to constitutively elevate HMGB1 levels or alter inflammatory responses, including autoimmune disorders, genetic metabolic diseases, hematological diseases, diabetes, chronic renal failure, or neoplasms. This exclusion criterion was implemented to minimize confounding variables and establish baseline HMGB1 performance in otherwise healthy children with sepsis, though we acknowledge this limits generalizability to these high-risk populations who frequently develop sepsis.

Case collection

A group of patients admitted to Xinxiang Medical University First Affiliated Hospital from March 2022 to December 2023, aged between 1 month and 18 years, was tracked in this prospective cohort study. The study included only those patients with a discharge diagnosis of severe sepsis and complete medical records, including comprehensive laboratory testing. Based on the inclusion and exclusion criteria, 46 children with severe sepsis were enrolled and categorized into shock and non-shock groups based on the presence or absence of shock.

Ethical review

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The research protocol was approved by the Ethics Committee of Xinxiang Medical University First Affiliated Hospital (ethical approval No. EC-022-175). All participants and their families provided informed consent and adhered to medical ethics guidelines.

Data collection

Baseline indicators collected included sex, age, height, weight, body temperature, and pre-hospital clinical course.

Laboratory indicators included coagulation function parameters, complete blood count indices, electrolyte levels, renal function indices, liver function markers, indices of humoral immune function, circulatory indices, cellular immune function markers, blood gas analysis indices, and clinical indices, such as the presence of dyspnea, liver and kidney dysfunction, brain injury, Glasgow Coma Scale score, pediatric Sequential Organ Failure Assessment (pSOFA) scores, Pediatric Critical Illness Score (PCIS) scores, number of involved organs, and infection sites.

Organ dysfunction definitions: organ dysfunction was defined according to pediatric-adapted criteria: (I) liver dysfunction: alanine aminotransferase (ALT) or aspartate aminotransferase (AST) greater than twice the upper limit of normal for age, or total bilirubin >2 mg/dL; (II) myocardial damage: cardiac troponin I >0.04 ng/mL or echocardiographic evidence of left ventricular ejection fraction <55%; (III) renal dysfunction: serum creatinine >1.5 times baseline value or urine output <0.5 mL/kg/hour for more than 6 hours; (IV) brain damage: Glasgow Coma Scale score <15 or new neurological deficit on clinical examination. These definitions were adapted from pediatric Sequential Organ Failure Assessment criteria and international consensus guidelines (12,13).

Specimen collection and measurement sepsis serum sample collection

Serum samples were collected from children suspected of sepsis within 24 h of admission following informed consent from their families. Specimen requirements: samples were collected in orange anticoagulant tubes to prevent hemolysis. Serum was separated and stored overnight at 2–8 ℃. Subsequently, samples were centrifuged at approximately 1,000 ×g for 20 min to harvest the supernatant, which was promptly tested or, if immediate testing was not feasible, stored at −80 ℃ to prevent repeated freeze-thaw cycles. Experimental Kit: the HMGB1 Enzyme-Linked Immunosorbent Assay (Elisa) Kit was sourced from Wuhan Fein Biotechnology Co. Ltd. (Wuhan, China).

Statistical analysis

Due to the limited sample size and non-normal distribution of certain variables, measurement data were expressed as median (M) and interquartile range (Q1, Q3). Missing data were almost nonexistent. Non-parametric Mann-Whitney U tests were used for intergroup comparisons, Chi-squared tests for categorical data, Spearman’s rank correlation for assessing correlations between variables, receiver operating characteristic (ROC) curves for diagnostic efficiency, and logistic regression analysis to determine risk factors for septic shock. Graph Prism software (version 8.0) was used for graphical representations and R 4.23 with RStudio (version 2023.12.1) for predictive model development. Statistical significance was set at P<0.05. Given the exploratory nature of this study and the inter-correlation among inflammatory markers, we did not apply Bonferroni correction for multiple comparisons. However, we acknowledge that findings with P values between 0.01 and 0.05 should be considered hypothesis-generating and require validation in independent cohorts.

Results

In this study, 46 children with severe sepsis were enrolled, of which 17 were complicated by shock and 29 were not. Univariate analysis demonstrated significant disparities in the pre-treatment levels of HMGB1, procalcitonin (PCT), serum amyloid A (SAA), interleukin-6 (IL-6), fibrin degradation products, D-dimer, serum calcium ions, urea, and complement C4 between the two groups (P<0.05). No significant differences were observed in height, weight, age, body temperature, C-reactive protein (CRP), pre-hospital course, some routine blood indices, serum potassium, sodium, chloride ions, total protein, ALT, AST, lactate dehydrogenase, immunoglobulins, complement C3, lactate, blood ammonia, T lymphocyte subsets, arterial blood pH, pSOFA score, PCIS score, and the number of organs involved in both groups (P>0.05). The shock group exhibited significantly higher levels of PCT (Z=−3.847, P<0.001), pretreatment HMGB1 (Z=−2.856, P=0.004), SAA (Z=−2.287, P=0.02), IL-6 (Z=−2.674, P=0.007), fibrin degradation products (Z=−2.117, P=0.03), D-dimer (Z=−2.311, P=0.02), and urea (Z=−2.492, P=0.01) than the non-shock group. Additionally, significantly lower levels of serum calcium ions (Z=−2.39, P=0.02) and complement C4 (Z=−2.277, P=0.02) were observed in the shock group (Figure 1 and Table 1). The most common infection sites were respiratory (n=28, 60.9%), followed by gastrointestinal (n=9, 19.6%), bloodstream (n=5, 10.9%), and other sites (n=4, 8.7%). No significant difference in infection site distribution was observed between the shock and non-shock groups (χ²=2.78, P=0.43). Microbiological cultures were positive in 36 patients (78.3%), with bacterial infections predominating.

Figure 1 Comparison of inflammatory indicators (A-E) and scores (F,G) between the shock group and the non-shock group. ns (not significant), P>0.05; *, P<0.05; **, P<0.01; ***, P<0.001. CRP, C-reactive protein; HMGB1, high mobility group box 1; IL-6, interleukin-6; PCIS, Pediatric Critical Illness Score; PCT, procalcitonin; pSOFA, pediatric Sequential Organ Failure Assessment; SAA, serum amyloid A.

The diagnostic efficacy of HMGB1, CRP, PCT, IL-6, pSOFA, and PCIS scores in predicting the onset of shock in children with severe sepsis was assessed using ROC curve analysis. Notably, pre-treatment levels of HMGB1, PCT, IL-6, and SAA were found to be statistically significant predictors, while CRP, pSOFA score, and PCIS score were not. The ROC curve with the highest area under the curve (AUC) belonged to PCT, with an AUC of 0.843. The optimal cutoff value was identified as 5.29 ng/mL, corresponding to a sensitivity and specificity of 88.24% and 65.52%, respectively. HMGB1, SAA, and IL-6 yielded AUCs ranging from 0.704 to 0.755 (Figure 2 and Table 2).

Figure 2 The reflection of HMGB1 and some scores on septic shock in children. CRP, C-reactive protein; HMGB1, high mobility group box 1; IL-6, interleukin-6; PCIS, Pediatric Critical Illness Score; PCT, procalcitonin; pSOFA, pediatric Sequential Organ Failure Assessment; SAA, serum amyloid A.

Univariate logistic regression analysis with the occurrence of septic shock as the dependent variable revealed that pre-treatment HMGB1, PCT, SAA, serum chloride, calcium ions, complement C4, and PCIS scores significantly correlated with the development of shock (P<0.05). Other variables did not demonstrate a significant association (P>0.05) (Table S1).

To ensure the precision and efficiency of variable selection, this study initially screened the independent variables through intergroup comparisons, univariate logistic regression, and clinical relevance. With the occurrence of septic shock as the dependent variable, a stepwise regression approach was employed in the multivariate logistic regression analysis, indicating that pretreatment HMGB1 and PCT levels were independent risk factors for the progression to septic shock in children with severe sepsis (P<0.05) (Table 3). Consequently, a predictive model was constructed and a nomogram was generated (Figure 3). The nomogram provides a user-friendly tool for individualized risk prediction. To calculate shock probability: (I) locate the patient’s HMGB1 value on the HMGB1 axis and draw a vertical line to the ‘Points’ axis to obtain the HMGB1 points; (II) repeat this process for the PCT value; (III) sum the points from both variables to obtain the total score; (IV) locate the total score on the ‘Total points’ axis and draw a vertical line down to the ‘Predicted probability’ axis to determine the shock risk. For example, a patient with HMGB1 of 1,591.74 pg/mL (corresponding to approximately 40 points) and PCT of 44 ng/mL (corresponding to approximately 35.5 points) would have a total score of 75.5 points, translating to a predicted shock probability of 73.2%.

Figure 3 Nomogram for predicting the development of shock in children with severe sepsis. The gray density map describes the distribution of various predictive factors and the total scores of children with severe sepsis included in the study. The top horizontal axis represents the score scale. When a variable takes different values, it corresponds to a different score. Finally, the total score is calculated. Predicted risk probability was obtained according to the total score. As shown in the figure: when HMGB1 takes 1,591.74 pg/mL and PCT takes 44 ng/mL, the corresponding total score is 75.5 points, indicating that the child has a high risk of septic shock, reaching 73.2%. *, P<0.05; **, P<0.01. HMGB1, high mobility group box 1; PCT, procalcitonin.

The model was internally validated using the bootstrap method, yielding a C-index of 0.869 (95% CI: 0.860–0.878). ROC curve analysis showed an AUC of 0.874 (0.757, 0.992) for the model, with a sensitivity and specificity of 82.4% and 89.7%, respectively, at a cut-off value of 0.398 (Figure 4). The calibration plot demonstrated that the calibration curve closely approximated the ideal curve, indicating satisfactory model fit (Figure 5). Furthermore, decision curve analysis suggested that within a threshold probability range of 20–99%, the decision curve of the predictive model was consistently above both the none and all lines, implying that the nomogram’s utilization in forecasting shock risk offered a higher net benefit, thereby confirming the model’s clinical applicability, as illustrated in Figure 6.

Figure 4 ROC curve of the prediction model. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

Figure 5 Calibration diagram of the prediction model.

Figure 6 Calibration curve of the prediction model. All lines indicate that all children with severe sepsis will develop septic shock, and none line indicates that all children with severe sepsis will not develop sepsis into shock.

Discussion

The significance of HMGB1 in pediatric sepsis research

Severe sepsis in children frequently triggers a systemic inflammatory cascade, leading to organ failure, expedited disease progression, increased therapeutic challenges, and potential deterioration of prognosis (14). HMGB1, a pivotal inflammatory mediator, is released after tissue injury and exacerbates inflammation via the circulatory system. Elevated serum HMGB1 levels in adult patients with sepsis correlate with disease severity and poorer outcomes (15), indicating that excessive HMGB1 secretion may contribute to disease exacerbation. To prevent disease progression and optimize pediatric sepsis outcomes, the assessment of serum HMGB1 levels is crucial. This study involved children with sepsis and a control group, comparing HMGB1 levels and analyzing their correlation with infection. This study evaluated the association between inflammatory markers and disease severity and prognosis (16).

The value of HMGB1 in assessing the development of septic shock in children with severe sepsis

Clinical characteristics of children with septic shock

Children with septic shock exhibited elevated pre-treatment serum HMGB1, SAA, PCT, IL-6, fibrin degradation products, D-dimer, and urea levels, along with decreased electrolyte calcium ions and complement C4 levels (17). Univariate Logistic regression identified pretreatment HMGB1, PCT, SAA, serum chloride, calcium ions, complement C4, and PCIS scores as associated with septic shock in severe pediatric sepsis. Multivariate regression revealed pretreatment PCT and HMGB1 levels as independent risk factors for septic shock, suggesting a link between increased HMGB1 levels and shock development. ROC analysis favored PCT over HMGB1 for assessment of septic shock. HMGB1 and PCT offer clinical predictive values for septic shock in severe pediatric sepsis (18).

The possible mechanism of HMGB1 in pediatric septic shock

Pediatric septic shock, a critical infectious condition, is characterized by systemic inflammation, hypotension, and multiple organ dysfunction. Its pathology involves a dysregulated host response to infection, leading to excessive cytokine production, widespread tissue injury, and organ failure (19). HMGB1, a nuclear protein released during cellular stress or damage, acts as a pro-inflammatory mediator, amplifying inflammation by activating TLR, interacting with RAGE, and modulating the NF-κB pathway (20). During septic shock, HMGB1 promotes cytokine release, induces vascular dysfunction, and contributes to organ dysfunction (21).

Pediatric septic shock and pSOFA score

The pSOFA score, an age-adapted version of the adult SOFA score, has been validated for quantifying organ dysfunction in critically ill children (22). In our study, while pSOFA scores showed a trend toward higher values in the shock group, this difference did not reach statistical significance. This may be attributed to our limited sample size and the fact that all enrolled patients had severe sepsis with substantial organ dysfunction, reducing the discriminative ability of pSOFA in this specific population. Nevertheless, pSOFA remains a valuable tool for overall severity assessment in pediatric sepsis. The latest pediatric sepsis consensus recommends the Phoenix score for diagnostic validation (23,24).

Pediatric septic shock and PCIS score

The PCIS score, a tool for assessing critical illness severity, is ineffective in pediatric septic shock assessment, likely due to its foundation of expert experience rather than large sample data and timely updates with clinical indicators (25). Recently, the Phoenix criteria have been proposed as a new international standard for pediatric sepsis diagnosis and stratification, with Phoenix Sepsis scores ≥2 indicating sepsis and Phoenix Septic Shock scores ≥1 indicating shock (26). These criteria were published in 2024, after our study enrollment period [2022–2023]. Our study used the 2015 Chinese Expert Consensus criteria, which were the institutional standard at the time. The relationship between HMGB1 levels and Phoenix scores represents an important area for future validation. We recommend that subsequent studies prospectively evaluate HMGB1’s predictive performance using Phoenix criteria in larger, multicenter cohorts, which would enhance the generalizability and contemporary relevance of our findings.

The study’s principal limitations include the following: (I) the analysis involved a small sample of 46 children with severe sepsis. The absence of a pre-determined sample size calculation may have obscured significant correlations or variances. Therefore, a larger sample size is warranted for future research; (II) the study’s temporal scope was confined to the measurement of HMGB1 levels within the initial 24 hours of hospital admission, omitting the continuous monitoring required to evaluate the association between HMGB1 levels and disease progression; (III) the study solely employed logistic regression analysis without assessing the predictive capabilities of alternative models. Subsequent research, with an expanded sample, aims to confirm these findings through machine learning methodologies; (IV) lastly, the study was conducted at a single institution, which precludes the inclusion of multi-center data for added validation; (V) we excluded patients with pre-existing conditions that could confound HMGB1 interpretation (autoimmune diseases, malignancies, etc.), which limits the generalizability to these high-risk populations who frequently develop sepsis. Future studies should specifically investigate HMGB1 utility in these complex patient populations; (VI) our study used the 2015 Chinese Expert Consensus for sepsis diagnosis, which was the institutional standard during the study period. While this represents a limitation given the subsequent publication of Sepsis-3 [2016] and Phoenix [2024] criteria, it reflects real-world practice evolution. External validation using contemporary diagnostic criteria is warranted.

Conclusions

Serum HMGB1 levels, particularly in combination with PCT, hold significant clinical value for early identification of septic shock risk in children with severe sepsis. Our predictive nomogram model (AUC 0.874) demonstrates strong discriminative ability and could serve as a practical bedside tool for risk stratification. However, external validation in larger, multicenter cohorts using contemporary diagnostic criteria (Sepsis-3 and Phoenix) is essential before clinical implementation. These findings support HMGB1 as a promising biomarker in pediatric sepsis management and highlight the need for further research to optimize its clinical utility.

Acknowledgments

We extend our heartfelt gratitude to the pediatric patients and their families for their participation and cooperation.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-703/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-703/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-703/prf

Funding: This study was supported by 2021 Henan Province Medical Science and Technology Research Plan (No. LHGJ20210519).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-aw-703/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The research protocol was approved by the Ethics Committee of Xinxiang Medical University First Affiliated Hospital (ethical approval No. EC-022-175). All participants and their families provided informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Zhang W, Jiang H, Wu G, et al. The pathogenesis and potential therapeutic targets in sepsis. MedComm (2020) 2023;4:e418.
2. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801-10. [Crossref] [PubMed]
3. Yang Z, Gao Y, Zhao L, et al. Molecular mechanisms of Sepsis attacking the immune system and solid organs. Front Med (Lausanne) 2024;11:1429370. [Crossref] [PubMed]
4. Li C, Qin C, Wei Y, et al. Interplays between host pattern-recognition receptors and pathogen ligands in immunogenic cell death. Microbes Immun 2024;3-19.
5. Zhu X, Huang H, Zhao L. PAMPs and DAMPs as the Bridge Between Periodontitis and Atherosclerosis: The Potential Therapeutic Targets. Front Cell Dev Biol 2022;10:856118. [Crossref] [PubMed]
6. Barichello T, Generoso JS, Singer M, et al. Biomarkers for sepsis: more than just fever and leukocytosis-a narrative review. Crit Care 2022;26:14. [Crossref] [PubMed]
7. Datta S, Rahman MA, Koka S, et al. High Mobility Group Box 1 (HMGB1): Molecular Signaling and Potential Therapeutic Strategies. Cells 2024;13:1946. [Crossref] [PubMed]
8. Kıral E, Yıldızdaş D, Ekinci F, et al. Serum Level of High-mobility Group Box Protein-1 in Children with Sepsis, Severe Sepsis and Septic Shock. J Pediatr Emerg Intensive Care Med 2024;11:174-8.
9. Paudel YN, Shaikh MF, Chakraborti A, et al. HMGB1: A Common Biomarker and Potential Target for TBI, Neuroinflammation, Epilepsy, and Cognitive Dysfunction. Front Neurosci 2018;12:628. [Crossref] [PubMed]
10. Wang X, Li R, Qian S, et al. Multilevel omics for the discovery of biomarkers in pediatric sepsis. Pediatr Investig 2023;7:277-89. [Crossref] [PubMed]
11. Subspecialty Group of Emergency Medicine. the Society of Pediatrics, Chinese Medical Association; Subspecialty Group of Pediatrics, the Society of Emergency Medicine, Chinese Medical Association; Pediatric Emergency Medicine Physicians, Chinese Medical Doctor Association. Zhonghua Er Ke Za Zhi 2015;53:576-80.
12. Goldstein B, Giroir B, Randolph A, et al. International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005;6:2-8. [Crossref] [PubMed]
13. Lim PPC, Bondarev DJ, Edwards AM, et al. The evolving value of older biomarkers in the clinical diagnosis of pediatric sepsis. Pediatr Res 2023;93:789-96. [Crossref] [PubMed]
14. He RR, Yue GL, Dong ML, et al. Sepsis Biomarkers: Advancements and Clinical Applications-A Narrative Review. Int J Mol Sci 2024;25:9010. [Crossref] [PubMed]
15. Jiang S, Liu L, Zhu X. Correlation of serum H-FABP, sTREM-1, and HMGB1 levels with severity and prognosis of sepsis. Am J Transl Res 2024;16:5846-55. [Crossref] [PubMed]
16. Fan Y, Ye Z, Tang Y. Serum HMGB1 and soluble urokinase plasminogen activator receptor levels aid diagnosis and prognosis prediction of sepsis with acute respiratory distress syndrome. Biomark Med 2023;17:231-9. [Crossref] [PubMed]
17. Zaki HA, Bensliman S, Bashir K, et al. Accuracy of procalcitonin for diagnosing sepsis in adult patients admitted to the emergency department: a systematic review and meta-analysis. Syst Rev 2024;13:37. [Crossref] [PubMed]
18. Xie Y, Zhuang D, Chen H, et al. 28-day sepsis mortality prediction model from combined serial interleukin-6, lactate, and procalcitonin measurements: a retrospective cohort study. Eur J Clin Microbiol Infect Dis 2023;42:77-85. [Crossref] [PubMed]
19. Ren W, Zhao L, Sun Y, et al. HMGB1 and Toll-like receptors: potential therapeutic targets in autoimmune diseases. Mol Med 2023;29:117. [Crossref] [PubMed]
20. Singh H, Agrawal DK. Therapeutic Potential of Targeting the HMGB1/RAGE Axis in Inflammatory Diseases. Molecules 2022;27:7311. [Crossref] [PubMed]
21. Li SJ, Ruan DD, Wu WZ, et al. Potential regulatory role of the Nrf2/HMGB1/TLR4/NF-κB signaling pathway in lupus nephritis. Pediatr Rheumatol Online J 2023;21:130. [Crossref] [PubMed]
22. García de Guadiana-Romualdo L, Botella LA, Rodríguez Rojas C, et al. Mortality prediction model from combined serial lactate, procalcitonin and calprotectin levels in critically ill patients with sepsis: A retrospective study according to Sepsis-3 definition. Med Intensiva (Engl Ed) 2024;48:629-38. [Crossref] [PubMed]
23. Mohamed El-Mashad G, Said El-Mekkawy M, Helmy Zayan M. Paediatric sequential organ failure assessment (pSOFA) score: A new mortality prediction score in the paediatric intensive care unit. An Pediatr (Engl Ed) 2020;92:277-85.
24. Schlapbach LJ, Straney L, Bellomo R, et al. Prognostic accuracy of age-adapted SOFA, SIRS, PELOD-2, and qSOFA for in-hospital mortality among children with suspected infection admitted to the intensive care unit. Intensive Care Med 2018;44:179-88. [Crossref] [PubMed]
25. Zhang Z, Huang X, Wang Y, et al. Performance of Three Mortality Prediction Scores and Evaluation of Important Determinants in Eight Pediatric Intensive Care Units in China. Front Pediatr 2020;8:522. [Crossref] [PubMed]
26. Sanchez-Pinto LN, Bennett TD, DeWitt PE, et al. Development and Validation of the Phoenix Criteria for Pediatric Sepsis and Septic Shock. JAMA 2024;331:675-86. [Crossref] [PubMed]