Diagnostic and prognostic value of neutrophil extracellular traps in sepsis and their correlation with inflammatory factors

Abstract

Objective

To investigate the plasma levels of neutrophil extracellular traps (NETs) biomarkers myeloperoxidase (MPO) and citrullinated histone H3 (CitH3) in sepsis patients, and to analyze their diagnostic value for sepsis, prognostic value for 28-day mortality, and correlation with inflammatory factors and organ function.

Methods

A total of 126 sepsis patients (divided into sepsis, severe sepsis and septic shock groups) and 60 healthy controls were enrolled. Sepsis patients were further divided into survival group and non-survival group according to 28-day prognosis. Plasma levels of MPO, CitH3, inflammatory factors (IL-6, IL-10, TNF-α, CXCL8) and clinical routine biomarkers (procalcitonin, PCT; lactate) were detected. Spearman correlation was used to analyze the relationship between NETs biomarkers and inflammatory factors, organ function indicators. Receiver operating characteristic (ROC) curve was used to evaluate the diagnostic efficacy of MPO and CitH3 for sepsis and their predictive value for 28-day mortality. Multivariate logistic regression was performed to identify independent predictors of 28-day mortality.

Results

Compared with healthy controls, plasma MPO, CitH3, IL-6, TNF-α, CXCL8 and PCT levels in sepsis patients were significantly increased, while IL-10 level was significantly decreased (all P < 0.001). MPO and CitH3 levels gradually increased with the aggravation of sepsis severity (septic shock > severe sepsis > sepsis, all P < 0.05). MPO and CitH3 were positively correlated with pro-inflammatory factors (IL-6, TNF-α, CXCL8), PCT and lactate, and negatively correlated with anti-inflammatory factor IL-10 (all P < 0.001). The levels of MPO and CitH3 in non-survival group were significantly higher than those in survival group (P < 0.001). ROC analysis showed that the area under curve (AUC) of MPO and CitH3 for sepsis diagnosis was 0.842 and 0.876, respectively, and the combined detection of MPO+CitH3 had a higher AUC of 0.931 (sensitivity 87.3%, specificity 93.3%). For 28-day mortality prediction, MPO+CitH3 achieved an AUC of 0.915, which was significantly higher than that of single index and PCT+lactate (P < 0.05).Multivariate logistic regression identified CitH3 (OR = 2.84, 95% CI: 1.62–4.98, P < 0.001), MPO (OR = 1.78, 95% CI: 1.21–2.62, P = 0.003), SOFA score (OR = 1.45, 95% CI: 1.12–1.88, P = 0.005), and lactate (OR = 1.32, 95% CI: 1.05–1.66, P = 0.018) as independent predictors of 28-day mortality.

Conclusion

Plasma MPO and CitH3 are highly expressed in sepsis patients, which are closely related to the severity of inflammation, organ dysfunction and short-term prognosis. Combined detection of MPO and CitH3 has high diagnostic value for sepsis and excellent prognostic efficacy for 28-day mortality, which can be used as potential biomarkers for clinical diagnosis, severity assessment and prognosis evaluation of sepsis.

Introduction

Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection, with high mortality in intensive care units worldwide [1]. Its pathogenesis involves excessive inflammatory activation, endothelial injury, and immune imbalance, leading to progressive organ failure and poor outcomes [2]. Early diagnosis, timely severity assessment, and accurate prognostic prediction are critical to improving clinical outcomes, yet traditional biomarkers such as procalcitonin (PCT) and lactate have limitations in comprehensive evaluation [3]. Thus, identifying novel biomarkers that integrate diagnostic and prognostic value remains an urgent clinical need.

Neutrophil extracellular traps (NETs) are web-like structures released by activated neutrophils, and their core components myeloperoxidase (MPO) and citrullinated histone H3 (CitH3) are specific biomarkers for NET formation [4]. Excessive NETs release aggravates systemic inflammation, vascular injury, and microthrombosis in sepsis, forming a positive feedback loop that amplifies inflammatory storm and organ damage [5]. Recent studies have confirmed that circulating MPO and CitH3 are elevated in sepsis and associated with disease severity, suggesting their potential as clinical indicators [6].

However, studies on the combined diagnostic and prognostic value of MPO and CitH3 in sepsis, as well as their systematic correlation with inflammatory factors, are still insufficient. In particular, the comparative performance of combined MPO and CitH3 detection versus traditional PCT and lactate combination for both sepsis diagnosis and 28-day mortality prediction has not been well established. Additionally, comprehensive correlation analysis between NETs biomarkers and multiple pro-inflammatory (IL-6, TNF-α, CXCL8) and anti-inflammatory (IL-10) cytokines in the same cohort remains limited.This study aimed to investigate plasma levels of MPO and CitH3 in sepsis patients, analyze their correlation with inflammatory factors and severity scores, and evaluate their efficacy in sepsis diagnosis and 28-day mortality prediction, to provide a new biomarker strategy for sepsis management.

Methods

Study design and participants

This prospective observational cohort study enrolled 126 sepsis patients and 60 healthy controls at our hospital from April 2023 to May 2025. Blood samples were collected prospectively upon enrollment (within 24 h of ICU admission and prior to antibiotic or vasopressor therapy), and clinical data including 28-day outcomes were recorded. Sepsis was diagnosed according to the Sepsis-3 criteria, and patients were further stratified into three subgroups based on disease severity for analytical purposes: (1) sepsis group: infection with SOFA score 2–5; (2) severe sepsis group: infection with SOFA score 6–10, not meeting septic shock criteria; and (3) septic shock group: sepsis with vasopressor requirement despite adequate fluid resuscitation and serum lactate > 2 mmol/L, per Sepsis-3 criteria [1]. Patients were also divided into survival and non-survival groups based on 28-day clinical outcomes.

Inclusion criteria for the sepsis group were: (1) age ≥ 18 years; (2) diagnosis of sepsis according to Sepsis-3 criteria (infection plus SOFA score ≥ 2 points); (3) admission to the intensive care unit (ICU); (4) availability of complete clinical and laboratory data. Exclusion criteria were: (1) age < 18 years; (2) pregnancy or lactation; (3) known immunocompromised status, including HIV infection, active hematological malignancy, solid organ transplantation, or long-term immunosuppressive therapy (glucocorticoids equivalent to prednisone ≥ 10 mg/day for > 3 months); (4) autoimmune diseases or chronic inflammatory disorders; (5) active malignancy requiring chemotherapy or radiotherapy within the past 6 months; (6) hospitalization for > 48 h prior to enrollment; (7) refusal to participate or lack of informed consent. Healthy controls were recruited from individuals undergoing routine health checkups at our hospital and were excluded if they had any acute or chronic infections, inflammatory diseases, or other significant medical conditions.

The study protocol was approved by the Ethics Committee of Linping first People’s Hospital (approval number: 2023044), and was performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants or their legal representatives prior to enrollment.

Data collection and laboratory measurements

Baseline demographic and clinical data including age, gender, BMI, SOFA score, and APACHE II score were recorded. Infection sources and microbiological findings were documented for all sepsis patients. Fasting peripheral venous blood was collected from all subjects in the early morning (after an 8-hour overnight fast) to minimize the confounding effects of postprandial changes on plasma volume and biomarker measurements, and to standardize collection timing across all participants. Plasma levels of MPO, CitH3, IL-6, TNF-α, IL-10, and CXCL8 were measured using enzyme-linked immunosorbent assay (ELISA) kits. PCT and lactate were detected by standard clinical biochemical methods. All experimental procedures were performed strictly in accordance with the manufacturer’s instructions.

Statistical analysis

All statistical analyses were performed using SPSS 26.0 and GraphPad Prism 7. Quantitative data were presented as mean ± standard deviation or median (interquartile range). Between-group comparisons were conducted using t-test, Mann-Whitney U test, or one-way ANOVA as appropriate. Spearman correlation analysis was used to evaluate the relationships between NETs biomarkers and inflammatory factors. Receiver operating characteristic (ROC) curves and Delong’s test were applied to assess the diagnostic and prognostic efficacy. The optimal cutoff values for MPO, CitH3, and other biomarkers were determined using the Youden index (maximum sensitivity + specificity – 1). For Kaplan-Meier survival analysis, patients were stratified into high and low groups based on the optimal cutoff value derived from the ROC curve for 28-day mortality prediction. Multivariate logistic regression (forward stepwise, likelihood ratio method) was performed to identify independent predictors of 28-day mortality; variables with P < 0.10 in univariate analysis were entered into the multivariate model, and results were presented as odds ratios (OR) with 95% confidence intervals (CI). A two-sided P < 0.05 was considered statistically significant.

Results

Baseline characteristics of study subjects

A total of 126 sepsis patients and 60 healthy controls were included. Baseline characteristics were well balanced among groups, with no significant differences in age, gender, or BMI (all P > 0.05). SOFA score, APACHE II score, and lactate level were significantly higher in the severe sepsis and septic shock subgroups than in the mild sepsis subgroup, and were markedly higher in non-survivors than in survivors (all P < 0.001) (Table 1).

Table 1.

Baseline characteristics of sepsis subgroups, survival and non-survival groups, and healthy controls

Characteristic	Healthy controls (n = 60)	Sepsis (n = 42)	Severe sepsis (n = 56)	Septic shock (n = 28)	Survival (n = 97)	Non-survival (n = 29)	F/χ²	P
Age (years)	45.6 ± 9.8	44.1 ± 11.2	46.3 ± 12.5	47.2 ± 13.6	45.2 ± 10.7	48.5 ± 12.3	0.612	0.608
Female, n (%)	35(58.3)	20(47.6)	27(48.2)	12(42.9)	47(48.5)	12(41.4)	1.184	0.757
BMI (kg/m²)	24.3 ± 3.2	24.7 ± 3.4	25.2 ± 3.8	25.7 ± 4.1	24.9 ± 3.5	26.1 ± 3.9	0.895	0.443
SOFA score	—	4.2 ± 1.3	7.8 ± 2.1	11.5 ± 2.6	6.1 ± 2.3	10.8 ± 2.8	132.54	< 0.001
APACHE II	—	12.6 ± 3.2	18.3 ± 4.1	24.7 ± 4.6	15.8 ± 4.2	23.6 ± 4.9	119.73	< 0.001
Lactate (mmol/L)	0.9 ± 0.3	2.1 ± 0.6	3.4 ± 0.9	5.2 ± 1.3	2.8 ± 0.8	4.9 ± 1.4	98.62	< 0.001
Infection source, n (%)
Pulmonary	—	18 (42.9)	26 (46.4)	14 (50.0)	45 (46.4)	13 (44.8)	—	—
Abdominal	—	10 (23.8)	15 (26.8)	7 (25.0)	24 (24.7)	8 (27.6)	—	—
Bloodstream	—	6 (14.3)	8 (14.3)	4 (14.3)	14 (14.4)	4 (13.8)	—	—
Urinary tract	—	5 (11.9)	4 (7.1)	3 (10.7)	9 (9.3)	3 (10.3)	—	—
Other	—	3 (7.1)	3 (5.4)	0 (0)	5 (5.2)	1 (3.4)	—	—
Microbiological confirmation, n (%)	—	34 (81.0)	45 (80.4)	23 (82.1)	78 (80.4)	24 (82.8)	—	—
Gram-negative	—	22 (64.7)	30 (66.7)	16 (69.6)	52 (66.7)	16 (66.7)	—	—
Gram-positive	—	8 (23.5)	9 (20.0)	5 (21.7)	16 (20.5)	6 (25.0)	—	—
Fungi	—	4 (11.8)	6 (13.3)	2 (8.7)	10 (12.8)	2 (8.3)	—	—

Note: Data are presented as mean ± standard deviation, n (%), or median (P25, P75). SOFA, Sequential Organ Failure Assessment; APACHE II, Acute Physiology and Chronic Health Evaluation II; BMI, body mass index

Among the 126 sepsis patients, the most common infection sources were: pulmonary infection (n = 58, 46.0%), abdominal infection (n = 32, 25.4%), bloodstream infection (n = 18, 14.3%), urinary tract infection (n = 12, 9.5%), and other sources (n = 6, 4.8%). Microbiological confirmation was obtained in 102 patients (81.0%), with the most frequently identified pathogens being: Gram-negative bacteria (n = 68, 66.7%), including Escherichia coli (n = 24), Klebsiella pneumoniae (n = 22), and Pseudomonas aeruginosa (n = 12); Gram-positive bacteria (n = 22, 21.6%), including Staphylococcus aureus (n = 14) and Streptococcus pneumoniae (n = 6); and fungi (n = 12, 11.8%), including Candida albicans (n = 8). No significant differences in MPO or CitH3 levels were observed across different infection sources or pathogen types (all P > 0.05).

Comparison of NETs biomarkers and inflammatory factors between sepsis and healthy controls

Compared with healthy controls, sepsis patients had significantly higher plasma levels of MPO, CitH3, IL-6, TNF-α, CXCL8, PCT, and lactate, while IL-10 was significantly lower (all P < 0.001) (Table 2). These results indicated that NETs formation and systemic inflammatory response were markedly activated in sepsis.

Table 2.

Comparison of NETs biomarkers, inflammatory factors, and classic sepsis indicators between sepsis group and healthy controls

Indicator	Sepsis group (n = 126)	Healthy controls (n = 60)	Z/t	P
MPO (ng/mL)	192.5 ± 46.8	89.8 ± 23.6	16.22	< 0.001
CitH3 (ng/mL)	4.12 ± 1.31	1.03 ± 0.32	19.45	< 0.001
IL-6 (pg/mL)	89.6 ± 24.5	10.9 ± 2.3	32.11	< 0.001
TNF-α (pg/mL)	42.8 ± 11.3	6.2 ± 1.5	28.66	< 0.001
IL-10 (pg/mL)	5.36 ± 1.24	18.9 ± 3.4	-29.88	< 0.001
CXCL8 (pg/mL)	112.4 ± 28.6	32.2 ± 10.5	21.35	< 0.001
PCT (ng/mL)	8.6(3.2,18.5)	0.05(0.03,0.09)	12.33	< 0.001
Lactate (mmol/L)	3.2(1.8,5.1)	0.9(0.7,1.1)	13.69	< 0.001

Data are presented as mean ± standard deviation or median (P25, P75). MPO, myeloperoxidase; CitH3, citrullinated histone H3; IL, interleukin; TNF-α, tumor necrosis factor-α; CXCL8, C-X-C motif chemokine ligand 8; PCT, procalcitonin

Expression of NETs biomarkers in different severity subgroups of sepsis

Plasma levels of MPO, CitH3, pro-inflammatory cytokines, PCT, and lactate increased gradually with sepsis severity (septic shock > severe sepsis > sepsis, all P < 0.05). In contrast, IL-10 showed a gradual decrease (all P < 0.05) (Table 3). These findings demonstrated that NETs biomarkers were closely associated with the severity of sepsis.

Table 3.

Comparison of NETs biomarkers, inflammatory factors, and classic indicators among different severity subgroups of sepsis

Indicator	Sepsis (n = 42)	Severe sepsis (n = 56)	Septic shock (n = 28)	F/Z	P
MPO	141.6 ± 33.5	196.8 ± 40.2ᵃ	272.4 ± 43.6ᵃᵇ	88.12	< 0.001
CitH3	2.56 ± 0.91	4.08 ± 1.12ᵃ	5.82 ± 1.24ᵃᵇ	94.36	< 0.001
IL-6	56.8 ± 15.2	92.4 ± 22.6ᵃ	136.8 ± 31.5ᵃᵇ	72.45	< 0.001
TNF-α	28.5 ± 7.2	43.6 ± 10.8ᵃ	62.8 ± 13.6ᵃᵇ	66.92	< 0.001
IL-10	6.58 ± 1.12	5.26 ± 1.08ᵃ	3.82 ± 0.71ᵃᵇ	22.16	< 0.001
CXCL8	78.6 ± 20.3	114.2 ± 26.8ᵃ	158.6 ± 32.4ᵃᵇ	81.37	< 0.001
PCT	3.8(1.6,7.2)	9.4(4.5,19.6)ᵃ	24.6(12.8,42.5)ᵃᵇ	79.55	< 0.001
Lactate	1.9(1.4,2.6)	3.5(2.4,4.8)ᵃ	5.4(4.1,7.2)ᵃᵇ	86.23	< 0.001

Data are presented as mean ± standard deviation or median (P25, P75). aP < 0.05 vs. sepsis subgroup; bP < 0.05 vs. severe sepsis subgroup. MPO, myeloperoxidase; CitH3, citrullinated histone H3; IL, interleukin; TNF-α, tumor necrosis factor-α; CXCL8, C-X-C motif chemokine ligand 8; PCT, procalcitonin

Differences in NETs biomarkers between survival and non-survival groups

Non-survivors exhibited significantly higher levels of MPO, CitH3, IL-6, TNF-α, PCT, and lactate, and lower IL-10 compared with survivors (all P < 0.001) (Table 4). This suggested that elevated NETs biomarkers were closely related to poor short-term prognosis in sepsis.

Table 4.

Comparison of NETs biomarkers and inflammatory factors between survival and non-survival groups in sepsis patients

Indicator	Survival (n = 97)	Non-survival (n = 29)	t/Z	P
MPO (ng/mL)	168.4 ± 38.5	246.8 ± 41.2	9.66	< 0.001
CitH3 (ng/mL)	3.56 ± 1.08	5.42 ± 1.16	8.72	< 0.001
IL-6 (pg/mL)	72.4 ± 18.6	128.6 ± 28.4	11.35	< 0.001
TNF-α (pg/mL)	36.8 ± 9.2	58.6 ± 12.4	10.11	< 0.001
IL-10 (pg/mL)	5.88 ± 1.16	4.12 ± 0.86	-6.92	< 0.001
PCT (ng/mL)	6.2(2.4,13.8)	22.8(11.5,39.6)	8.66	< 0.001
Lactate (mmol/L)	2.6(1.6,4.2)	5.2(3.8,7.4)	9.33	< 0.001

Data are presented as mean ± standard deviation or median (P25, P75). MPO, myeloperoxidase; CitH3, citrullinated histone H3; IL, interleukin; TNF-α, tumor necrosis factor-α; CXCL8, C-X-C motif chemokine ligand 8; PCT, procalcitonin

Correlation between NETs biomarkers and inflammatory factors

Spearman correlation analysis showed that MPO and CitH3 were significantly positively correlated with IL-6, TNF-α, and CXCL8, and negatively correlated with IL-10 (all P < 0.001) (Fig. 1). These results supported that NETs were involved in the inflammatory imbalance of sepsis.

Fig. 1.

Correlation between CitH3 and IL-6 in sepsis patients

Diagnostic efficacy of NETs biomarkers for sepsis

ROC curve analysis revealed that the AUC of MPO and CitH3 for sepsis diagnosis was 0.842 and 0.876, respectively. The combined detection of MPO + CitH3 achieved a higher AUC of 0.931 (sensitivity 87.3%, specificity 93.3%), which was superior to single markers and the PCT + lactate combination (P < 0.05) (Table 5; Fig. 2).

Table 5.

ROC curve results of different biomarkers for the diagnosis of sepsis

Indicator	AUC	95%CI	Sensitivity	Specificity	P
MPO	0.842	0.778–0.906	80.2%	86.7%	< 0.001
CitH3	0.876	0.817–0.935	83.3%	88.3%	< 0.001
MPO+CitH3	0.931	0.892–0.970	87.3%	93.3%	< 0.001
PCT+lactate	0.884	0.826–0.942	84.1%	86.7%	< 0.001

AUC, area under the receiver operating characteristic curve; CI, confidence interval; MPO, myeloperoxidase; CitH3, citrullinated histone H3; PCT, procalcitonin

Fig. 2.

ROC curves of NETs biomarkers for the diagnosis of sepsis

Prognostic value of NETs biomarkers for 28-day mortality

For predicting 28-day mortality, the AUC of MPO + CitH3 was 0.915, which was significantly higher than that of single biomarkers and PCT + lactate (P < 0.05) (Table 6; Fig. 3). The optimal cutoff values for predicting 28-day mortality were: MPO > 215.6 ng/mL and CitH3 > 4.58 ng/mL. Kaplan-Meier survival analysis showed that patients with high CitH3 levels (defined as > 4.58 ng/mL, the optimal cutoff from ROC analysis for 28-day mortality prediction) had a significantly lower cumulative survival rate compared to those with low CitH3 levels (log-rank P < 0.001) (Fig. 4).

Table 6.

ROC curve results of different biomarkers for predicting 28-day mortality in sepsis patients

Indicator	AUC	95%CI	Sensitivity	Specificity	P
MPO	0.816	0.735–0.897	79.3%	82.5%	< 0.001
CitH3	0.858	0.784–0.932	82.8%	86.6%	< 0.001
MPO+CitH3	0.915	0.861–0.969	89.7%	91.8%	< 0.001
PCT+lactate	0.863	0.790–0.936	83.5%	85.6%	< 0.001

AUC, area under the receiver operating characteristic curve; CI, confidence interval; MPO, myeloperoxidase; CitH3, citrullinated histone H3; PCT, procalcitonin

Fig. 3.

ROC curves of NETs biomarkers for predicting 28-day mortality in sepsis patients

Fig. 4.

Kaplan-Meier survival curves of sepsis patients grouped by CitH3 level

Multivariate logistic regression analysis was performed to identify independent predictors of 28-day mortality. Variables with P < 0.10 in univariate analysis (age, SOFA score, APACHE II score, lactate, PCT, MPO, CitH3, IL-6, TNF-α, and IL-10) were entered into the multivariate model. After adjustment for these potential confounders, CitH3 (OR = 2.84, 95% CI: 1.62–4.98, P < 0.001), MPO (OR = 1.78, 95% CI: 1.21–2.62, P = 0.003), SOFA score (OR = 1.45, 95% CI: 1.12–1.88, P = 0.005), and lactate (OR = 1.32, 95% CI: 1.05–1.66, P = 0.018) remained independent predictors of 28-day mortality (Table 7). These findings confirm that MPO and CitH3 provide independent prognostic information beyond traditional clinical severity scores and biomarkers.

Table 7.

Multivariate logistic regression analysis of independent predictors for 28-day mortality

Variable	β	SE	Wald χ²	OR	95% CI	P
CitH3 (per 1 ng/mL increase)	1.044	0.287	13.24	2.84	1.62–4.98	< 0.001
MPO (per 10 ng/mL increase)	0.577	0.195	8.75	1.78	1.21–2.62	0.003
SOFA score (per 1 point increase)	0.372	0.133	7.82	1.45	1.12–1.88	0.005
Lactate (per 1 mmol/L increase)	0.278	0.118	5.55	1.32	1.05–1.66	0.018

OR, odds ratio; CI, confidence interval; SE, standard error; MPO, myeloperoxidase; CitH3, citrullinated histone H3; SOFA, Sequential Organ Failure Assessment. Variables entered into the multivariate model: age, SOFA score, APACHE II score, lactate, PCT, MPO, CitH3, IL-6, TNF-α, and IL-10. Hosmer-Lemeshow goodness-of-fit test: χ² = 6.82, P = 0.556. Nagelkerke R² = 0.423

Discussion

Sepsis is a life-threatening clinical syndrome defined by dysregulated host response to infection, which remains a leading cause of death in intensive care units worldwide [1]. Its complex pathogenesis involves excessive inflammatory activation, immune dysfunction, endothelial damage, and progressive organ impairment, making early diagnosis and risk stratification extremely challenging [2]. Although procalcitonin (PCT) and lactate are widely used in clinical practice, their limitations in comprehensive evaluation have prompted the search for more effective biomarkers [3].

Neutrophil extracellular traps (NETs) are web-like structures released by activated neutrophils, which participate in antimicrobial defense but also promote inflammatory injury in critical illness [4]. Excessive NET formation aggravates systemic inflammation, microthrombosis, and tissue damage, thereby contributing to the progression of sepsis [5]. As specific biomarkers for NETs, myeloperoxidase (MPO) and citrullinated histone H3 (CitH3) have been increasingly recognized as promising indicators for assessing disease activity and severity in sepsis [6].

In the present study, plasma MPO and CitH3 levels were significantly higher in sepsis patients than in healthy controls and gradually increased with disease severity, suggesting that NETs activation is closely associated with the progression of sepsis. These results are consistent with previous reports indicating that circulating NETs markers are elevated in septic patients and correlate with organ dysfunction scores, and are closely associated with the degree of inflammatory response [7]. Our study extends previous findings by demonstrating that combined MPO and CitH3 detection not only distinguishes sepsis from healthy controls with superior diagnostic accuracy compared to PCT + lactate (AUC 0.931 vs. 0.884), but also provides excellent prognostic performance for 28-day mortality (AUC 0.915). Furthermore, the systematic correlation analysis with multiple inflammatory cytokines in the same patient cohort provides novel insights into the mechanistic role of NETs in sepsis-related inflammatory imbalance [8].

It should be noted that the direction of IL-10 change in sepsis varies across studies, with some reports showing elevated IL-10 levels in sepsis patients compared to healthy controls [9]. This heterogeneity likely reflects differences in patient populations, timing of sample collection, and disease stage. In our cohort, patients were enrolled at the time of sepsis diagnosis, corresponding to the early hyperinflammatory phase, which may explain the relatively lower IL-10 levels observed. Furthermore, the progressive decrease in IL-10 with increasing disease severity suggests that more severe sepsis is associated with greater immune dysregulation.

Notably, MPO and CitH3 levels were significantly higher in non-survivors than in survivors, and the combined MPO + CitH3 panel showed excellent performance in predicting 28-day mortality, with an AUC of 0.915. Multivariate logistic regression further confirmed that MPO and CitH3 were independent predictors of 28-day mortality after adjusting for age, SOFA score, APACHE II score, lactate, PCT, and inflammatory cytokines, with CitH3 showing the strongest association (OR = 2.84, 95% CI: 1.62–4.98, P < 0.001). This finding supports the role of NETs as prognostic biomarkers in sepsis, which is in line with recent clinical studies [10]. Compared with traditional biomarkers such as PCT and lactate, the combined detection of MPO and CitH3 provides higher sensitivity and specificity, which may help clinicians identify high-risk patients at an early stage [11].

Previous studies have also investigated the relationship between NETs and inflammatory regulation in sepsis. For instance, increased NETs formation has been shown to promote systemic inflammation and contribute to multiple organ dysfunction through the release of pro-inflammatory mediators [12]. Additionally, the correlation between NETs biomarkers and cytokine profiles observed in our study is supported by experimental evidence demonstrating that NETs components can activate macrophages and endothelial cells, thereby amplifying the inflammatory cascade [13]. Furthermore, recent studies have elucidated the molecular mechanisms underlying NETosis and inflammatory activation in sepsis, providing a biological basis for our clinical findings [14]. The role of NETs in sepsis-induced coagulation and vascular injury has also been well documented [15]. Moreover, CitH3 has been specifically identified as an early biomarker of NETosis in septic organ damage, which is consistent with our observation that CitH3 was the strongest independent predictor of 28-day mortality among all biomarkers evaluated [16].

This study has several limitations. First, it is a single-center study with a relatively limited sample size. Second, we only detected plasma levels of NETs markers without measuring local tissue expression. Third, the specific mechanism by which NETs regulate inflammation in sepsis was not fully explored. Fourth, as an observational study, we cannot establish causality between elevated NETs biomarkers and adverse outcomes. Future prospective, multi-center studies with larger sample sizes are needed to verify these results.

Conclusion

In conclusion, plasma MPO and CitH3 are closely related to inflammation, disease severity, and short-term prognosis in sepsis. Combined detection of these two NETs biomarkers provides high diagnostic and prognostic value, which can serve as a novel tool for the clinical management of sepsis.

Abbreviations

NETs

Neutrophil extracellular traps

MPO

Myeloperoxidase

CitH3

Citrullinated histone H3

IL

Interleukin

TNF-α

Tumor necrosis factor-α

CXCL8

CXC motif chemokine ligand 8

PCT

Procalcitonin

SOFA

Sequential Organ Failure Assessment

APACHE II

Acute Physiology and Chronic Health Evaluation II

Author contributions

Xueyan Zhou: Conceptualization, supervision, project administration, writing – review & editing, corresponding author. Zhaofeng Bao: Data curation, formal analysis, writing – original draft. Junling Zhou, Zhangwei Xia: Data collection, investigation, methodology, manuscript revision. All authors have read and approved the final manuscript.

Funding

This research received no external funding.

Data availability

The raw data supporting the conclusions of this article are provided in the supplementary material. All individual data points, baseline characteristics, and laboratory measurements are included in the Excel file. The data can be made available upon request from the corresponding author without undue reservation.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Ethics Committee of Linping Campus, The Second Affiliated Hospital of Zhejiang University (approval number: 2023044), and was performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants or their legal representatives prior to enrollment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.