Next Article in Journal
Weight Reduction with GLP-1 Agonists and Paths for Discontinuation While Maintaining Weight Loss
Previous Article in Journal
Electrostatic Effects on Tau Nanocondensates
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Biomolecules
Volume 15
Issue 3
10.3390/biom15030407
biomolecules-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Systematic Review

IL-6 Baseline Values and Dynamic Changes in Predicting Sepsis Mortality: A Systematic Review and Meta-Analysis

by
Norberth-Istvan Varga
1,
Iulia Cristina Bagiu
2,3,*,
Dan Dumitru Vulcanescu
2,3,
Voichita Lazureanu
4,
Mirela Turaiche
4,5,
Ovidiu Rosca
4,5,
Adrian Vasile Bota
6 and
Florin George Horhat
2,3
1
Department of General Medicine, Doctoral School, “Victor Babes” University of Medicine and Pharmacy, Eftimie Murgu Square No. 2, 300041 Timisoara, Romania
2
Department of Microbiology, “Victor Babes” University of Medicine and Pharmacy, Eftimie Murgu Square No. 2, 300041 Timisoara, Romania
3
Multidisciplinary Research Center on Antimicrobial Resistance (MULTI-REZ), Microbiology Department, “Victor Babes” University of Medicine and Pharmacy, Eftimie Murgu Square No. 2, 300041 Timisoara, Romania
4
Department XIII, Discipline of Infectious Diseases, “Victor Babes” University of Medicine and Pharmacy, Eftimie Murgu Square No. 2, 300041 Timisoara, Romania
5
Methodological and Infectious Diseases Research Center, Department of Infectious Diseases, “Victor Babes” University of Medicine and Pharmacy, Eftimie Murgu Square No. 2, 300041 Timisoara, Romania
6
Doctoral School, Faculty of Medicine, “Vasile Goldis” Western University, Bulevardul Revolutiei 94, 310025 Arad, Romania
*
Author to whom correspondence should be addressed.
Biomolecules 2025, 15(3), 407; https://doi.org/10.3390/biom15030407
Submission received: 16 January 2025 / Revised: 4 March 2025 / Accepted: 12 March 2025 / Published: 13 March 2025
(This article belongs to the Section Molecular Medicine)
Browse Figures
Versions Notes

Abstract

:
Sepsis, a life-threatening condition arising from a dysregulated immune response to infection, is a significant health burden globally. Interleukin-6 (IL-6), an inflammatory cytokine produced by immune cells as a response to infection and tissue damage, plays a key role in the pathogenesis of sepsis. This systematic review and meta-analysis aimed to investigate the association of the baseline plasma levels of IL-6, and the dynamic change in these levels over a timespan of 96 h, with short-term mortality. A systematic literature search was conducted across multiple databases. Studies were included if they assessed the independent prognostic value of IL-6 in adult sepsis patients, used well-defined sepsis criteria, and reported at least one IL-6 measurement. Pooled effect estimates for the association between IL-6 and 28–30-day mortality were determined using logistic regression and AUROC analysis. Thirty-one studies, encompassing 4566 patients, were included. While baseline IL-6 levels and 96 h IL-6 clearance were not significantly associated with mortality risk (pooled OR 1.001, 95% CI 0.999–1.003 and 1.019, 95% CI 0.925–1.112, respectively), AUROC analysis indicated moderate-to-good discriminatory power for both baseline (0.701, 95% CI 0.660–0.742) and 96 h IL-6 clearance (0.828, 95% CI 0.736–0.919) in predicting 28-day mortality. While not a strong independent predictor, IL-6 demonstrates some discriminatory ability, suggesting its potential value in conjunction with other biomarkers.

1. Introduction

Sepsis, a life-threatening condition characterized by a dysregulated immune response to infection, represents a major global public health burden [1,2]. With an estimated 49 million cases and 11 million deaths annually, sepsis accounts for nearly 20% of all global deaths [2]. An initial infectious trigger activates a dysregulated immune response, characterized by a potentially fatal cascade of inflammatory mediators [3]. These mediators initiate a complex and dynamic interplay between pro- and anti-inflammatory processes [4]. While the pro-inflammatory response aims to eradicate the invading pathogen, an excessive reaction can lead to collateral tissue damage. A compensatory anti-inflammatory process that also includes various cytokines is set into motion, but it can also be inadequate or exaggerated. This tug-of-war involves not just IL-6, but a chorus of markers, like tumor necrosis factor-alpha (TNF-α), which kickstarts the cytokine storm, and C-reactive protein (CRP), a downstream signal of inflammation, all shaping sepsis’s chaotic course [5]. Sepsis is characterized by the disruption of this delicate balance between pro-inflammatory and anti-inflammatory mechanisms. The systemic inflammatory state that results is marked by widespread endothelial dysfunction, microvascular abnormalities, such as increased permeability, vasodilation, microthrombi formation, and, ultimately, impaired tissue perfusion and organ dysfunction [6,7,8].
Interleukins are a type of cytokine produced by the immune cells (T-helper cells, monocytes and macrophages, B cells, etc.), but also by endothelial and epithelial cells, as a response to infection or injury, and play a central role in the intricate inflammatory network that characterizes sepsis [9]. Their main roles include immune cell development and differentiation, immune cell activation and proliferation, and regulation of the immune response. This includes pro-inflammatory interleukins like IL-1β and IL-8, which amplify the early attack on pathogens, and anti-inflammatory ones like IL-10, which temper the response, alongside IL-6, within a broader family that includes leukemia inhibitory factor (LIF) and oncostatin M (OSM). Interleukin 6 (IL-6) stimulates hepatic synthesis of acute-phase proteins, such as C-reactive protein and fibrinogen [10,11]. It also plays an important role in adaptive immunity by promoting the differentiation of B cells into antibody-secreting plasma cells and influencing the development of various T cell subsects, including pro-inflammatory Th17 cells [10,11,12]. Systemically, IL-6 induces fever and contributes to muscle wasting. In sepsis, IL-6 levels are markedly elevated, due to its increased production by a variety of immune cells: monocytes, macrophages, and endothelial cells [13]. This surge in IL-6 plasma values, coupled with its early appearance in the course of the disease, reflects its utility as a biomarker for sepsis [14].
Building upon the established diagnostic utility of IL-6 in sepsis, the primary focus of this study shifts towards elucidating its prognostic value. Far from acting alone, IL-6 interacts with many other mediators, making its prognostic role trickier to pin down. While elevated IL-6 levels are clearly associated with the presence and severity of sepsis, a more nuanced understanding of their relationship with patient outcomes is needed. Therefore, this systematic review and meta-analysis aims to comprehensively evaluate the available evidence, and to answer pivotal questions regarding the prognostic significance of IL-6 in sepsis. Specifically, it seeks to determine whether higher baseline IL-6 levels are independently associated with an increased risk of mortality in septic patients. Furthermore, it investigates whether the dynamic changes in IL-6 concentrations over time, i.e., IL-6 kinetics, provide additional prognostic information beyond static baseline measurements, as well as determining their ability to predict sepsis mortality. By synthesizing data from original articles reporting serial IL-6 measurements in sepsis, this meta-analysis aims to provide a definitive assessment of IL-6’s role as a prognostic biomarker and its potential to guide clinical decision-making in the management of this life-threatening condition.

2. Materials and Methods

2.1. Search Strategy

This systematic review and meta-analysis adhered to established best practices. For methodological transparency, the protocol was prospectively registered (PROSPERO ID CRD42024515819), and is reported in accordance with PRISMA guidelines [15]. A comprehensive search strategy across Google Scholar, Cochrane Library, and Pubmed employed the following keywords: “IL-6”, “interleukin-6”, “interleukin-6 in sepsis”, “IL-6 in sepsis”, “sepsis prognosis”, “sepsis prognostic factor”, “sepsis biomarkers”, and “sepsis mortality”. To capture the most relevant advancements, our focus prioritized publications from the past two decades (January 2000–December 2023). The review included original research articles with the following study designs: prospective and retrospective observational cohort studies, cross-sectional studies, and case–control studies.

2.2. Study Selection

Two independent researchers separately reviewed titles, abstracts, and full-text articles to select papers eligible for our study. If discrepancies arose, a third reviewer resolved those disagreements. Only studies focusing on the prognostic factor’s independent ability to predict patient outcomes were chosen, i.e., studies that measured baseline IL-6 values and/or IL-6 kinetics. The inclusion criteria were as follows: (1) studies that explicitly measured the biomarker’s independent predictive power related to a well-defined outcome (in-hospital mortality or 28–30-day mortality). Only phase 2 confirmatory studies were included (phase 2 studies explore “the independence of the association between a prognostic factor and the outcome of interest … while controlling for confounders” [16]); (2) studies with more than 10 patients and five outcome events; (3) studies only on adult patients (18 years or older); (4) studies that used well-defined criteria for sepsis diagnosis; and (5) studies that reported at least one measurement of the prognostic factor (baseline values and/or kinetics of IL-6 levels). The study selection process, following the PRISMA guidelines, is illustrated in Figure 1.
The exclusion criteria were as follows: studies primarily reporting associations without strong statistical analysis or without controlling for confounders (phase 1 studies), case series, studies unrelated to our primary focus, and studies on patients suffering from immunocompromised conditions.

2.3. Data Analysis

Two separate reviewers independently extracted the relevant data from the included studies, and systematically organized them into a table, ensuring a methodical approach for our analysis. The following data were extracted: study ID (first author and year of publication), country, clinical setting, number of participants, outcome events, mean or median age of participants, sex (percentage of males or females), criteria used for sepsis definition, sepsis origin, clinical scores (SOFA, APACHE-II), plasma IL-6 values, logistic regression analysis, area under the receiver operating characteristics (AUROC) curves to determine outcome, covariates, etc. Some studies used different units of measurement for IL-6, but we transformed these units into pg/mL for clarity. Studies in which IL-6 was transformed into a logarithmic scale were analyzed separately, to appropriately account for methodological differences. A transformation back to a common unit would complicate direct comparisons and pooling of results.
Data on plasma IL-6 values at hospital admission and subsequent measurements were carefully noted and analyzed, along with their associated timings (e.g., 24, 48 h). We prioritized corresponding effect measures from multivariable models, focusing on adjusted odds ratios (ORs) and hazard ratios (HRs). These OR and HR values were often reported at multiple timepoints throughout the studies, and any provided area under the receiver operating characteristics (AUROC) curves to determine outcomes related to baseline and follow-up intervals were captured.
The primary outcome of this meta-analysis was mortality, specifically, all-cause mortality, sepsis-related mortality, and in-hospital mortality, with a particular focus on the 28–30-day timeframe. This widely accepted clinical endpoint captures both early deaths associated with the acute phase of sepsis, and those occurring due to later complications or persistent organ dysfunction. All included studies reported mortality data either as in-hospital mortality or within the specified 28–30-day follow-up period, allowing for standardized assessment of short-term outcomes and the prognostic value of interleukin-6 (IL-6). Based on this outcome, patients were classified into two groups: survivors (S) and non-survivors (NS), facilitating analysis of IL-6’s potential impact on survival.

2.4. Risk of Bias Assessment

We assessed the risk of bias within our included studies using the Quality In Prognosis Studies (QUIPS) tool [17], tailored specifically for our review (Supplemental Materials, Tables S2 and S3). The QUIPS tool evaluates six key domains: study participation (D1), study attrition (D2), prognostic factor measurement (D3), outcome measurement (D4), confounding measurement (D5), and statistical analysis and reporting (D6). Two reviewers independently scored each study for risk of bias across all domains, resolving discrepancies through discussion with a third expert. As suggested by the tool guidelines, each study was then assigned a “low”, “moderate”, or “high” risk of bias rating for each domain. In the confounding domain (D5), while acknowledging that different studies adjusted for diverse covariates (most often other biomarkers like procalcitonin, lactate, WBC, severity scores, body mass index, alcohol dependence, cancer, etc.), our expert group focused on age and at least one severity score (SOFA or APACHE-II) as critical and consistent adjustment factors for this review.

2.5. Statistical Analysis

To evaluate the association between baseline IL-6 levels and 28-day mortality, we performed two types of meta-analyses. The first type was a meta-analysis of effect measures: we pooled the odds ratios (ORs), hazard ratios (HRs), and risk ratios (RRs) reported in the included studies. Where available, we prioritized effect measures derived from multivariable models that adjusted for potential confounders. The second type was a meta-analysis of the area under the receiver operating characteristics (AUROC) curve: we pooled the AUROC values to assess the discriminatory ability of baseline IL-6 levels for predicting 28-day mortality.
For IL-6 kinetics within the first 96 h after admission, we also conducted meta-analyses of both effect measures and AUROC values, using the same approach as described above. In all meta-analyses, we employed a random effects restricted maximum likelihood (REML) model to account for heterogeneity between studies. Pooled estimates were calculated using the Sidik–Jonkman method with Hartung–Knapp 95% confidence intervals. Heterogeneity was assessed using the I2 statistic and Cochran’s Q test. Publication bias was evaluated using Egger’s and Begg’s tests.
All meta-analyses were performed using MedCalc software for Windows, version 23.1.7 (MedCalc Software, Ostend, Belgium).

3. Results

3.1. Overview of Selected Studies

The initial search of the electronic databases, using the aforementioned keywords, resulted in 22,337 studies in total. An initial examination of titles led to the exclusion of 8149 duplicates and 12,931 studies with other research aims (not evaluating the prognostic value of IL-6 in sepsis). After a thorough examination of the abstracts of the remaining studies, a total of 1208 studies were excluded because they did not serve our study objectives (n = 743), used unclear definitions for sepsis (n = 48), or employed other study designs (n = 407), such as reviews, case reports, etc. The remaining articles were assessed for eligibility through full-text review. The studies that were excluded [17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33] from the meta-analysis after the final screening step, and the reasons for exclusion, are described in the Supplemental Materials, Table S1.
Our selection process resulted in a final set of 31 studies [34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64], including a total of 4566 septic patients. Out of these 31 studies, 16 investigated the fluctuations in subsequent (i.e., post-baseline) IL-6 measurements, and the corresponding measure of effect on the outcome. All 31 studies used mortality as the primary outcome, with slight variations: 25 studies examined 28–30-day mortality periods, while 4 focused on in-hospital mortality without specific timelines. One study reported outcomes at 4-day, 28-day, and 1-year timepoints, but only the first 28 days were accounted for in the meta-analysis [30]. Publication trends revealed that 3 articles were published between 2000 and 2010, 17 between 2010 and 2020, and 11 between 2020 and 2023. Most studies were performed in China, encompassing 10 of the included studies, followed by Spain, with 4, and South Korea, Greece, and Japan, with 2. Of the included studies, 24 were prospective, with 7 utilizing a retrospective design. The clinical settings in which the studies were performed were mostly ICU wards (22 studies), while 5 studies were performed in the ER, 3 studies were performed in mixed clinical settings (ICU + a medical ward), and 1 study was performed in the Department of Traumatology and Acute Critical Medicine. The definitions of sepsis differed greatly across studies, reflecting their publication dates: 18 used the Singer 2015 criteria, 10 employed the Levy 2001 criteria, and 3 used other well-defined criteria. A general overview of the selected studies is provided in Table 1.
The number of patients included in the studies ranged from 29 to 708, with a median of 93 (25th–75th percentiles 62.5–174). From a total of 4566 patients, the percentage of men was 62.58% on average, ranging across studies from 43.8% to 81.8%. Age reporting varied across studies, with all providing either mean or median values. In 12 studies, age was reported separately for the two groups of patients, survivors (S) vs. non-survivors (NS). Twenty-six studies reported a mixed origin of sepsis in their patients, one article focused exclusively on patients with sepsis of respiratory origin, and four studies did not report the origin of sepsis.

3.2. Risk of Bias

Across all studies, the QUIPS domain with the highest risk of bias was statistical analysis and reporting (domain 6), for which 15/31 studies received a “high” rating. The reasons for this were as follows: reporting exclusively on statistically significant values, limitations in describing statistical methodologies, or underutilizing available data (for instance, some articles failed to report information such as the area under the receiver operating characteristics curve to determine outcomes, IL-6 cut-off values, sensibility, specificity, positive and negative predictive values, etc.). If two or more items were considered problematic, a “high” rating was given. Seven articles received a “moderate” rating, reflecting only one of the aforementioned concerns, and nine studies received a “low” rating. Confounding measurements (Domain 5) presented challenges in assessment because of the wide variation in prognostic confounders across studies. The two most frequent covariates were age and at least one severity score (APACHE-II or SOFA), justifying our decision to use these as the main covariates. Studies failing to adjust for both at a minimum received a “high” risk of bias rating in Domain 5. Overall, the remaining domains of the QUIPS tool posed a low-to-moderate risk of bias across the majority of the studies. A complete breakdown of our tailored application of the QUIPS domains, along with the rationale used for every assessment, is provided in the Supplemental Materials, Tables S2 and S3.

3.3. Meta-Analysis

A comprehensive data extraction table was constructed for this study, documenting study ID, SOFA and APACHE-II scores, baseline IL-6 levels, IL-6 clearance at various timepoints, and effect estimates from IL-6 logistic regression and AUROC analyses (with 95% confidence intervals). All the data that were gathered from each individual article and used for the meta-analysis can be found in Table S4.
As illustrated in Table 2, among the 31 studies, 12 reported the association between IL-6 levels and mortality using odds ratios (ORs), with a total of 2004 patients. Five studies reported hazard ratios (HRs), encompassing 461 patients. One study with 708 patients reported risk ratios (RRs). Twelve studies did not report suitable effect estimates for inclusion in the meta-analysis. Two studies employed log-transformed ORs, including data from 284 patients.
Regarding model adjustments, 13 studies accounted for both age and disease severity, with 7 utilizing the SOFA score, 7 using APACHE-II, and 5 incorporating both. Age was the most frequently adjusted-for confounder (15 studies), followed by SOFA (12 studies) and APACHE-II (11 studies). Nineteen studies included adjustments for various biomarkers. Notably, studies adjusting for both age and severity tended to report higher odds ratios (median OR 1.13) compared to those adjusting for age alone (median OR 1.033) or not adjusting for age (median OR 1.002), suggesting that these factors may confound the relationship between IL-6 and mortality.

3.3.1. Baseline IL-6 Values and IL-6 Clearance: Effect Measures

The overall pooled odds ratio at baseline (OR) is 1.001, suggesting that higher baseline IL-6 levels are not associated with a significant increase in mortality risk. The 95% confidence interval (CI) for the pooled OR is very narrow, ranging from 0.999 to 1.003. This further supports the conclusion of no statistically significant association between baseline IL-6 levels and mortality. The forest plot can be observed in Figure 2, while the results of the meta-analysis can be seen in Table 3.
The test for heterogeneity (Q = 31.8881, p = 0.0042) reveals significant heterogeneity between the studies included in the meta-analysis. This suggests that factors other than chance contribute to the differences in results across studies, such as variations in study design, patient populations, or IL-6 measurement methods. The I2 statistic (56.10%) confirms moderate heterogeneity. Egger’s test shows a significant p-value (p = 0.0005), suggesting potential publication bias. This means that studies with positive or significant results might be more likely to be published, leading to an overestimation of the true effect in the meta-analysis. Begg’s test is also significant (p = 0.0234), further supporting the presence of publication bias.
Regarding IL-6 clearance in the 96 h timeframe, the overall pooled OR is 1.019, indicating that elevated IL-6 levels in the first 96 h after admission are not associated with a statistically significant increase in mortality risk. The 95% confidence interval (CI) for the pooled OR ranges from 0.925 to 1.112. Since this interval includes 1, we cannot be confident that the true effect is different from no effect. The results of the meta-analysis can be found in Table 4, with the respective forest plot in Figure 3.
The test for heterogeneity (Q = 2.7022, p = 0.2590) reveals no significant heterogeneity between the three studies included in the meta-analysis. This implies that the observed differences in results across studies are likely due to chance, rather than systematic differences in study design or population. The I2 statistic (25.99%) indicates low heterogeneity. Egger’s test shows a non-significant p-value (p = 0.3970), suggesting no significant evidence of publication bias based on this test. Begg’s test is also not significant (p = 0.1172), further supporting the absence of publication bias. Based on these results, we cannot conclude that IL-6 levels within the first 96 h after admission have significant prognostic value for predicting mortality in sepsis patients using logistic regression. Only three studies could be used to perform this test, which limits the power to detect a small but clinically significant effect.
In a separate analysis of two studies reporting log-transformed odds ratios (284 patients), a significant association was found between elevated IL-6 levels and mortality in sepsis patients. At admission, a one-unit increase in log IL-6 was associated with an 81% increase in the odds of mortality (pooled log OR 1.81, 95% CI 1.44–2.27). This association persisted on day 3, with a one-unit increase in log IL-6 corresponding to a 165% increase in the odds of mortality (pooled log OR 2.65, 95% CI 1.99–3.53). These findings suggest that IL-6 may be a valuable prognostic marker in this subset of studies, although the generalizability of these results to the broader population of sepsis patients is limited by the small number of studies included in this analysis and the methodological differences in reporting effect estimates. Due to the inclusion of only two studies, formal assessment of heterogeneity and publication bias was not conducted.

3.3.2. Area Under the Receiver Operating Characteristic Curve

The overall pooled AUROC is 0.701 (95% CI: 0.660–0.742), indicating that the ability of baseline IL-6 levels to predict 28-day mortality in sepsis patients is moderate. The test for heterogeneity (Q = 8874.2114, p < 0.0001) reveals significant heterogeneity between the studies. This means that the differences in AUROC values between studies are likely due to factors other than chance, such as differences in study design, patient populations, or IL-6 measurement methods. The I2 statistic (99.88%) further confirms the substantial heterogeneity. The results of the statistical testing can be found in Table 5, with the respective forest plot in Figure 4.
Egger’s test shows a non-significant p-value (p = 0.4378), indicating no significant evidence of publication bias based on this test. Begg’s test is also not significant (p = 0.4929), further supporting the absence of publication bias. Based on these results, baseline IL-6 appears to have a moderate ability to predict mortality in sepsis patients. However, the significant heterogeneity between studies suggests that the predictive performance of IL-6 may vary depending on various factors.
Regarding IL-6 clearance, the overall pooled AUROC is 0.828 (95% CI: 0.736–0.919), indicating that IL-6 levels within the first 96 h have good discriminatory power for predicting mortality in sepsis patients. The test for heterogeneity (Q = 1.4741, p = 0.2247) shows no significant heterogeneity between the two studies included in the meta-analysis. This implies that the observed differences in results across studies are likely due to chance, rather than systematic differences in study design or population. The I2 statistic (32.16%) with a confidence interval spanning from 0.00 to 100.00 also reflects the uncertainty due to the small number of studies. The results of the statistical testing can be found in Table 6, with the respective forest plot in Figure 5.
Egger’s test reveals a significant p-value (<0.0001), suggesting potential publication bias. This means that studies with positive or significant results might be more likely to be published, leading to an overestimation of the true effect in the meta-analysis. Begg’s test, however, is not significant (p = 0.3173), indicating no significant evidence of publication bias based on this test alone. These results imply that IL-6 levels within the first 96 h after admission appear to have good potential as a predictor of mortality in sepsis patients. However, the confidence interval for the pooled AUROC is relatively wide (0.736 to 0.919), which suggests that there is still considerable uncertainty regarding the true discriminatory power of IL-6, due to the limited number of studies included.

4. Discussion

This systematic review and meta-analysis sought to clarify the prognostic value of interleukin-6 (IL6) in predicting mortality among critically ill patients with sepsis. Our investigation, encompassing both baseline IL-6 levels and their dynamic changes over time, revealed a nuanced picture. While traditional effect measures like odds ratios (ORs) and hazard ratios (HRs) did not demonstrate a statistically significant association between IL-6 and mortality, analyses of the area under the receiver operating characteristic curve (AUROC) unveiled the cytokine’s potential as a discriminatory tool. This suggests that IL-6, though perhaps not a potent independent predictor of mortality, may, nonetheless, possess value in differentiating between patients at higher and lower risk of death, a finding with potential implications for risk stratification and personalized therapeutic strategies.
Our observation that baseline IL-6 levels are not significantly associated with increased mortality risk aligns with the findings of Molano-Franco et al. [65], who similarly reported no significant association between the baseline values of several biomarkers, including IL-6, and mortality in septic patients. However, our study differs in that it focused exclusively on IL-6, and incorporated a larger number of studies (n = 31) and patients (n = 4566) than the Molano-Franco et al. analysis. Additionally, our work uniquely examined the prognostic value of IL-6 kinetics, analyzing changes in IL-6 levels over the first 96 h, whereas the Molano-Franco et al. study primarily evaluated baseline values. This expanded scope offers a more comprehensive assessment of IL-6’s potential role as a prognostic biomarker in sepsis. In contrast to the Li et al. [66] meta-analysis, which examined cytokine changes after treatment initiation, our study specifically examined the prognostic potential of IL-6 dynamics within the first 96 h, independently of therapeutic interventions. These findings, along with our own, highlight the complexities of using IL-6 as a single prognostic indicator in sepsis, and suggest the need for a more nuanced approach.
The concept of biomarker clearance as a predictor of outcomes in sepsis patients has been explored with other inflammatory markers as well, particularly procalcitonin (PCT), C-Reactive Protein (CRP), lactate (LAC), etc. Researchers have sought to determine whether a significant drop in the plasma levels of these markers, within a certain timeframe, is associated with better outcomes. For instance, Charles et al. (2009) found that a PCT decrease of >30% before the third day of hospitalization was an independent predictor of survival in ICU patients with sepsis [67]. Similarly, Karlsson et al. (2010) reported lower mortality in patients with a >50% decrease in PCT by 72 h [68]. Other studies have also demonstrated the prognostic value of PCT and CRP clearance, with varying definitions and timeframes [69,70,71,72,73].
The divergence between our findings based on pooled effect measures, such as odds ratios (ORs) and hazard ratios (HRs), and those derived from the area under the receiver operating characteristic curve (AUROC) merits further exploration. Odds ratios and hazard ratios estimate how strongly IL-6 levels predict mortality risk—OR reflects the odds of death with higher IL-6, while HR tracks risk over time—yet our results showed no significant link, suggesting that any risk increase is minor or negligible across a broad patient group. A key complicating factor is the wide variability in IL-6 levels observed in sepsis. In our clinical studies, IL-6 concentrations spanned a broad range, which could have reduced the reliability of statistical tests like OR and HR by introducing inconsistency that diluted their ability to detect a clear pattern. Even in preclinical research with tightly controlled conditions, such as the mouse model study by Lee et al. (2022), IL-6 levels fluctuated significantly following an inflammatory trigger, indicating that this variability is an inherent biological trait rather than a measurement artifact [74]. This natural fluctuation likely weakened our pooled OR and HR estimates, explaining their lack of statistical significance despite IL-6’s moderate predictive strength in AUROC analyses. In contrast, the AUROC assesses how well IL-6 distinguishes between patients who survive and those who do not, regardless of the exact risk magnitude. Our moderate-to-good AUROC values for baseline and early IL-6 measurements suggest that it can still flag higher-risk patients effectively. This contrast highlights a critical point: a lack of statistical significance in risk estimates does not necessarily diminish clinical utility. Though IL-6 may not sharply elevate mortality odds, its capacity to stratify risk could still inform patient management decisions.
Our findings are complicated by the considerable heterogeneity we observed across the studies we analyzed. This variation stems from a multitude of factors related to the patient populations of our included studies: the severity of sepsis, IL-6 measurement techniques, and statistical methodologies. Notably, the variability in sepsis definitions across the included studies, ranging from older criteria (Levy 2001) [75] to more recent ones (Singer 2015) [1], could introduce heterogeneity in the assessment and prognosis of sepsis. Our findings must also be considered in the context of aging, which significantly shapes IL-6 dynamics. According to Ershler and Keller [76], aging drives elevated baseline IL-6 levels through chronic low-grade inflammation—termed “inflammaging”—potentially reducing its specificity as a prognostic marker in older septic patients. Similarly, Ferrucci et al. [77] highlighted that this persistent IL-6 elevation in the elderly correlates with frailty and immune dysregulation, which could mask acute sepsis-related spikes. A study by Puzianowska-Kuźnicka et al. [78] further demonstrated that age-related shifts in IL-6 receptor signaling, especially in T cells, impair immune responses, possibly contributing to the heterogeneity we observed across studies with diverse median ages. These insights emphasize the need for age-stratified analyses in future research to unravel IL-6’s prognostic role across the lifespan.
Furthermore, the potential for publication bias, as indicated by some of our analyses, warrants consideration. While we have strived to include a comprehensive range of studies, it is possible that studies with positive or significant findings were more likely to be published, potentially leading to an overestimation of the true effect. This observation emphasizes the necessity for careful interpretation of data, and highlights the importance of future research that addresses these limitations.
Beyond age, genetic variability adds another layer to IL-6’s prognostic complexity. Single-nucleotide polymorphisms (SNPs) in genes like IL-10, an anti-inflammatory cytokine, can shift its activity [79,80]. Overactive variants might dampen IL-6’s inflammatory drive, while underactive ones could let it run unchecked, altering plasma levels and their link to mortality. Our meta-analysis, built on summary data, could not tease out these effects, but they likely fueled some of the heterogeneity we saw across studies. This suggests that genetic profiles could muddy IL-6’s signal, a wrinkle worth exploring in future work. IL-6 itself carries genetic variability. As a member of a cytokine family steering acute-phase responses, IL-6’s gene on chromosome 7p21 harbors variants like the −174 C/G promoter polymorphism [28]. The GG genotype boosts IL-6 production over the CC, potentially intensifying inflammation and mortality risk in sepsis [28,81]. Without genotype-specific data, our analysis could not probe this, but it is a plausible driver of IL-6’s uneven prognostic strength, highlighting the urgency of future genotype-focused studies.
Our study has several noteworthy limitations. Firstly, despite including a substantial number of studies that investigated the role of IL-6 in sepsis prognosis, there was a limited number of studies that reported the dynamic changes in plasma IL-6 over time. The significant heterogeneity observed across the included studies also introduces a degree of uncertainty into our conclusions. These factors limit the generalizability of our findings. Secondly, our focus on 28–30-day mortality, while reflecting the predominant trend in sepsis research, may have inadvertently excluded valuable insights from studies examining longer-term outcomes, such as 90-day mortality. These longer-term outcomes, however, might be significantly influenced by factors that are not related to sepsis (e.g., old age, comorbidities, complications, etc.). These confounding factors can make it challenging to isolate the specific contribution of any specific biomarker to mortality risk over extended periods. Thirdly, the generalizability of our findings might also be affected by the potential for bias in the primary studies, variability in patient characteristics, and the exclusion of certain populations (patients under the age of 18, or immunocompromised individuals). Lastly, relying on a single biomarker like IL-6 to predict mortality in a complex syndrome like sepsis has inherent limitations. Sepsis is a multifaceted phenomenon, driven by a complex interplay of pathophysiological processes, and a single molecule does not adequately capture this complexity. Even if our findings suggest a limited independent prognostic value for IL-6, our systematic review and meta-analysis provides future research suggestions and prevents redundant efforts in the future.
Moving forward, research should prioritize addressing the limitations identified in this review and exploring new avenues for understanding the role of IL-6 in sepsis prognosis. Future studies should delve deeper into the sources of heterogeneity observed in IL-6 studies, potentially employing techniques like meta-regression to analyze the relationship between study characteristics and effect sizes. Given the limitations of single biomarkers, the development and validation of multi-marker panels that incorporate IL-6 alongside other promising candidates, such as procalcitonin, CRP, and lactate, hold immense potential for improving prognostic accuracy. This could be further enhanced by leveraging machine learning techniques to identify optimal biomarker combinations and generate more precise risk stratification models. To address the limitations of relying on summary data, future research should prioritize collecting and analyzing individual patient data. This would allow for a more granular investigation of IL-6 kinetics, including the determination of optimal IL-6 thresholds for predicting mortality and the exploration of potential interactions with other biomarkers. Furthermore, dedicated studies are needed to investigate the prognostic value of IL-6 in specific sepsis subpopulations, such as immunosuppressed patients, pediatric patients, and those with different infection sources or pathogens.

5. Conclusions

In conclusion, while our systematic review and meta-analysis suggest that IL-6, whether measured at baseline or within the first 96 h after admission, may not be a reliable independent predictor of mortality in critically ill patients with sepsis, the moderate-to-good discriminatory ability of IL-6 observed in the AUROC analyses warrants further investigation. Future research should focus on identifying sources of heterogeneity, exploring the impact of different IL-6 thresholds, and considering the use of IL-6 as part of a multi-marker panel, or in specific patient subgroups, to unlock its potential in refining sepsis prognosis and guiding personalized treatment decisions.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/biom15030407/s1: Table S1: Excluded studies and reasons for exclusion; Table S2: QUIPS domains for risk of bias assessment; Table S3: Risk of bias ratings; Table S4: Full dataset used for meta-analysis.

Author Contributions

Conceptualization, N.-I.V. and F.G.H.; methodology, I.C.B.; validation, I.C.B. and F.G.H.; formal analysis, D.D.V. and M.T.; investigation, N.-I.V.; data curation, N.-I.V. and A.V.B.; writing—original draft preparation, O.R. and M.T.; writing—review and editing, V.L.; visualization, F.G.H.; supervision, I.C.B.; project administration, F.G.H. All authors have read and agreed to the published version of the manuscript.

Funding

The APC was covered by the “Victor Babes” University of Medicine and Pharmacy, Timisoara.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

We would like to acknowledge the “Victor Babes” University of Medicine and Pharmacy, Timisoara, Romania, for its support in covering the costs of publication for this research paper.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Singer, M.; Deutschman, C.S.; Seymour, C.W.; Shankar-Hari, M.; Annane, D.; Bauer, M.; Bellomo, R.; Bernard, G.R.; Chiche, J.; Coopersmith, C.M.; et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016, 315, 801–810. [Google Scholar] [CrossRef] [PubMed]
  2. Rudd, K.E.; Johnson, S.C.; Agesa, K.M.; Shackelford, K.A.; Tsoi, D.; Kievlan, D.R.; Colombara, D.V.; Ikuta, K.S.; Kissoon, N.; Finfer, S.; et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: Analysis for the Global Burden of Disease Study. Lancet 2020, 395, 200–211. [Google Scholar] [CrossRef]
  3. Chousterman, B.G.; Swirski, F.K.; Weber, G.F. Cytokine storm and sepsis disease pathogenesis. Semin. Immunopathol. 2017, 39, 517–528. [Google Scholar] [CrossRef]
  4. Clark, I.A.; Vissel, B. The meteorology of cytokine storms, and the clinical usefulness of this knowledge. Semin. Immunopathol. 2017, 39, 505–516. [Google Scholar] [CrossRef] [PubMed]
  5. Chaudhry, H.; Zhou, J.; Zhong, Y.; Ali, M.M.; McGuire, F.; Nagarkatti, P.S.; Nagarkatti, M. Role of cytokines as a double-edged sword in sepsis. Vivo 2013, 27, 669–684. [Google Scholar] [PubMed]
  6. Jacobi, J. Pathophysiology of sepsis. Am. J. Health Pharm. 2002, 59 (Suppl. S1), S3–S8. [Google Scholar] [CrossRef]
  7. Schouten, M.; Wiersinga, W.J.; Levi, M.; van der Poll, T. Inflammation, endothelium, and coagulation in sepsis. J. Leukoc. Biol. 2008, 83, 536–545. [Google Scholar] [CrossRef]
  8. Levi, M.; van der Poll, T. Coagulation and sepsis. Thromb. Res. 2017, 149, 38–44. [Google Scholar] [CrossRef]
  9. Akdis, M.; Aab, A.; Altunbulakli, C.; Azkur, K.; Costa, R.A.; Crameri, R.; Duan, S.; Eiwegger, T.; Eljaszewicz, A.; Ferstl, R.; et al. Interleukins (from IL-1 to IL-38), interferons, transforming growth factor β, and TNF-α: Receptors, functions, and roles in diseases. J. Allergy Clin. Immunol. 2016, 138, 984–1010. [Google Scholar] [CrossRef]
  10. Uciechowski, P.; Dempke, W.C. Interleukin-6: A Masterplayer in the Cytokine Network. Oncology 2020, 98, 131–137. [Google Scholar] [CrossRef]
  11. Wolf, J.; Rose-John, S.; Garbers, C. Interleukin-6 and its receptors: A highly regulated and dynamic system. Cytokine 2014, 70, 11–20. [Google Scholar] [CrossRef] [PubMed]
  12. Mihara, M.; Hashizume, M.; Yoshida, H.; Suzuki, M.; Shiina, M. IL-6/IL-6 receptor system and its role in physiological and pathological conditions. Clin. Sci. 2011, 122, 143–159. [Google Scholar] [CrossRef] [PubMed]
  13. Rose-John, S. IL-6 trans-signaling via the soluble il-6 receptor: Importance for the pro-inflammatory activities of IL-6. Int. J. Biol. Sci. 2012, 8, 1237–1247. [Google Scholar] [CrossRef]
  14. Hou, T.; Huang, D.; Zeng, R.; Ye, Z.; Zhang, Y. Accuracy of serum interleukin (IL)-6 in sepsis diagnosis: A systematic review and meta-analysis. Int. J. Clin. Exp. Med. 2015, 8, 15238–15245. [Google Scholar] [PubMed]
  15. Hayden, J.A.; Côté, P.; Steenstra, I.A.; Bombardier, C.; QUIPS-LBP Working Group. Identifying phases of investigation helps planning, appraising, and applying the results of explanatory prognosis studies. J. Clin. Epidemiol. 2008, 61, 552. [Google Scholar] [CrossRef]
  16. Hayden, J.A.; Van Der Windt, D.A.; Cartwright, J.L.; Côté, P.; Bombardier, C. Assessing bias in studies of prognostic factors. Ann. Intern. Med. 2013, 158, 280–286. [Google Scholar] [CrossRef]
  17. Amancio, R.T.; Japiassu, A.M.; Gomes, R.N.; Mesquita, E.C.; Assis, E.F.; Medeiros, D.M.; Grinsztejn, B.; Bozza, P.T.; Castro-Faria, H.C.; Bozza, F.A. The innate immune response in hiv/aids septic shock patients: A comparative study. PLoS ONE 2013, 8, e68730. [Google Scholar] [CrossRef]
  18. Viallon, A.; Guyomarc’h, S.; Marjollet, O.; Berger, C.; Carricajo, A.; Robert, F.; Laporte, S.; Lambert, C.; Page, Y.; Zéni, F.; et al. Can emergency physicians identify a high mortality subgroup of patients with sepsis: Role of procalcitonin. Eur. J. Emerg. Med. 2008, 15, 26–33. [Google Scholar] [CrossRef]
  19. Calandra, T.; Gerain, J.; Heumann, D.; Baumgartner, J.-D.; Glauser, M.P. High circulating levels of interleukin-6 in patients with septic shock: Evolution during sepsis, prognostic value, and interplay with other cytokines. Am. J. Med. 1991, 91, 23–29. [Google Scholar] [CrossRef]
  20. Kandaswamy, P.; Hemlata; Singh, G.P.; Ahmad, M.K. Comparative Evaluation of Procalcitonin and Interleukin-6 as Diagnostic and Prognostic Biomarkers for Sepsis. J. Clin. Diagn. Res. 2018, 12, UC17–UC21. [Google Scholar] [CrossRef]
  21. Kellum, J.A.; Kong, L.; Fink, M.P.; Weissfeld, L.A.; Yealy, D.M.; Pinsky, M.R.; Fine, J.; Krichevsky, A.; Delude, R.L.; Angus, D.C. Understanding the inflammatory cytokine response in pneumonia and sepsis: Results of the genetic and inflammatory markers of sepsis (GenIMS) study. Arch. Intern. Med. 2007, 167, 1655–1663. [Google Scholar] [CrossRef] [PubMed]
  22. Tschaikowsky, K.; Hedwig-Geissing, M.; Braun, G.G.; Radespiel-Troeger, M. Predictive value of procalcitonin, interleukin-6, and C-reactive protein for survival in postoperative patients with severe sepsis. J. Crit. Care 2011, 26, 54–64. [Google Scholar] [CrossRef] [PubMed]
  23. Wunder, C.; Eichelbrönner, O.; Roewer, N. Are IL-6, IL-10 and PCT plasma concentrations reliable for outcome prediction in severe sepsis? A comparison with APACHE III and SAPS II. Inflamm. Res. 2004, 53, 158–163. [Google Scholar] [CrossRef]
  24. Jahn, M.; Rekowski, J.; Jánosi, R.A.; Kribben, A.; Canbay, A.; Katsounas, A. Score performance of SAPS 2 and SAPS 3 in combination with biomarkers IL-6, PCT or CRP. PLoS ONE 2020, 15, e0238587. [Google Scholar] [CrossRef]
  25. Ebihara, T.; Matsumoto, H.; Matsubara, T.; Togami, Y.; Nakao, S.; Matsuura, H.; Kojima, T.; Sugihara, F.; Okuzaki, D.; Hirata, H.; et al. Cytokine Elevation in Severe COVID-19 From Longitudinal Proteomics Analysis: Comparison with Sepsis. Front. Immunol. 2022, 12, 798338. [Google Scholar] [CrossRef]
  26. Jekarl, D.W.; Kim, J.Y.; Ha, J.H.; Lee, S.; Yoo, J.; Kim, M.; Kim, Y. Diagnosis and Prognosis of Sepsis Based on Use of Cytokines, Chemokines, and Growth Factors. Dis. Markers 2019, 1, 1089107. [Google Scholar] [CrossRef] [PubMed]
  27. Davoudian, S.; Piovani, D.; Desai, A.; Mapelli, S.N.; Leone, R.; Sironi, M.; Valentino, S.; Silva-Gomes, R.; Stravalaci, M.; Asgari, F.; et al. A cytokine/PTX3 prognostic index as a predictor of mortality in sepsis. Front. Immunol. 2022, 13, 979232. [Google Scholar] [CrossRef]
  28. Lorente, L.; Martín, M.M.; Pérez-Cejas, A.; Barrios, Y.; Solé-Violán, J.; Ferreres, J.; Labarta, L.; Díaz, C.; Jiménez, A. Association between Interleukin-6 Promoter Polymorphism (-174 G/C), Serum Interleukin-6 Levels and Mortality in Severe Septic Patients. Int. J. Mol. Sci. 2016, 17, 1861. [Google Scholar] [CrossRef]
  29. Frimpong, A.; Owusu, E.D.A.; Amponsah, J.A.; Obeng-Aboagye, E.; van der Puije, W.; Frempong, A.F.; Kusi, K.A.; Ofori, M.F. Cytokines as Potential Biomarkers for Differential Diagnosis of Sepsis and Other Non-Septic Disease Conditions. Front. Cell. Infect. Microbiol. 2022, 12, 901433. [Google Scholar] [CrossRef]
  30. Gentile, L.F.; Cuenca, A.G.; Vanzant, E.L.; Efron, P.A.; McKinley, B.; Moore, F.; Moldawer, L.L. Is there value in plasma cytokine measurements in patients with severe trauma and sepsis? Methods 2013, 61, 3–9. [Google Scholar] [CrossRef]
  31. Harbarth, S.; Holeckova, K.; Froidevaux, C.; Pittet, D.; Ricou, B.; Grau, G.E.; Vadas, L.; Pugin, J.; The Geneva Sepsis Network. Diagnostic value of procalcitonin, interleukin-6, and interleukin-8 in critically ill patients admitted with suspected sepsis. Am. J. Respir. Crit. Care Med. 2001, 164, 396–402. [Google Scholar] [CrossRef] [PubMed]
  32. Jekarl, D.W.; Kim, J.Y.; Lee, S.; Kim, M.; Kim, Y.; Han, K.; Woo, S.H.; Lee, W.J. Diagnosis and evaluation of severity of sepsis via the use of biomarkers and profiles of 13 cytokines: A multiplex analysis. Clin. Chem. Lab. Med. 2015, 53, 575–581. [Google Scholar] [CrossRef] [PubMed]
  33. Oda, S.; Hirasawa, H.; Shiga, H.; Nakanishi, K.; Matsuda, K.-I.; Nakamua, M. Sequential measurement of IL-6 blood levels in patients with systemic inflammatory response syndrome (SIRS)/sepsis. Cytokine 2005, 29, 169–175. [Google Scholar] [CrossRef] [PubMed]
  34. Andaluz-Ojeda, D.; Bobillo, F.; Iglesias, V.; Almansa, R.; Rico, L.; Gandía, F.; Resino, S.; Tamayo, E.; de Lejarazu, R.O.; Bermejo-Martin, J.F. A combined score of pro- and anti-inflammatory interleukins improves mortality prediction in severe sepsis. Cytokine 2012, 57, 332–336. [Google Scholar] [CrossRef]
  35. Belli, O.E.; Campolo, J.; Vallerio, P.; Musca, F.; Moreo, A.; Maloberti, A.; Parolini, M.; Bonacchini, L.; Monti, G.; De Gasperi, A.; et al. Biochemical but not imaging parameters are predictive of outcome in septic shock: A pilot study. Cardiovasc. Ultrasound 2022, 20, 6. [Google Scholar] [CrossRef]
  36. Beneyto, L.A.P.; Luis, O.R.; Sánchez, C.S.; Simón, O.C.; Rentero, D.B.; Bayarri, V.M. Valor pronóstico de la interleucina 6 en la mortalidad de pacientes con sepsis. Med. Clin. 2016, 147, 281–286, Erratum in Med. Clin. 2017, 148, 433. [Google Scholar] [CrossRef]
  37. Eidt, M.V.; Nunes, F.B.; Pedrazza, L.; Caeran, G.; Pellegrin, G.; Melo, D.A.; Possuelo, L.; Jost, R.T.; Dias, H.B.; Donadio, M.V.; et al. Biochemical and inflammatory aspects in patients with severe sepsis and septic shock: The predictive role of IL-18 in mortality. Clin. Chim. Acta 2016, 453, 100–106. [Google Scholar] [CrossRef]
  38. Frencken, J.F.; van Vught, L.A.; Peelen, L.M.; Ong, D.S.Y.; Klouwenberg, P.M.C.K.; Horn, J.; Bonten, M.J.M.; van der Poll, T.; Cremer, O.L. An Unbalanced Inflammatory Cytokine Response Is Not Associated with Mortality Following Sepsis: A Prospective Cohort Study. Crit. Care Med. 2017, 45, e493–e499. [Google Scholar] [CrossRef] [PubMed]
  39. Jekarl, D.W.; Lee, S.-Y.; Lee, J.; Park, Y.-J.; Kim, Y.; Park, J.H.; Wee, J.H.; Choi, S.P. Procalcitonin as a diagnostic marker and IL-6 as a prognostic marker for sepsis. Diagn. Microbiol. Infect. Dis. 2013, 75, 342–347. [Google Scholar] [CrossRef]
  40. Jiang, Y.; Jiang, F.-Q.; Kong, F.; An, M.-M.; Jin, B.-B.; Cao, D.; Gong, P. Inflammatory anemia-associated parameters are related to 28-day mortality in patients with sepsis admitted to the ICU: A preliminary observational study. Ann. Intensiv. Care 2019, 9, 67. [Google Scholar] [CrossRef]
  41. Karamouzos, V.; Giamarellos-Bourboulis, E.J.; Velissaris, D.; Gkavogianni, T.; Gogos, C. Cytokine production and outcome in MDR versus non-MDR gram-negative bacteraemia and sepsis. Infect. Dis. 2021, 53, 764–771. [Google Scholar] [CrossRef] [PubMed]
  42. Karampela, I.; Christodoulatos, G.S.; Kandri, E.; Antonakos, G.; Vogiatzakis, E.; Dimopoulos, G.; Armaganidis, A.; Dalamaga, M. Circulating eNampt and resistin as a proinflammatory duet predicting independently mortality in critically ill patients with sepsis: A prospective observational study. Cytokine 2019, 119, 62–70. [Google Scholar] [CrossRef] [PubMed]
  43. Liu, S.; Wang, X.; She, F.; Zhang, W.; Liu, H.; Zhao, X. Effects of Neutrophil-to-Lymphocyte Ratio Combined with Interleukin-6 in Predicting 28-Day Mortality in Patients with Sepsis. Front. Immunol. 2021, 12, 639735. [Google Scholar] [CrossRef]
  44. Liu, J.; Bai, C.; Li, B.; Shan, A.; Shi, F.; Yao, C.; Zhang, Y.; Wang, J.; Chen, W.; Xie, M.; et al. Mortality prediction using a novel combination of biomarkers in the first day of sepsis in intensive care units. Sci. Rep. 2021, 11, 1275. [Google Scholar] [CrossRef]
  45. Matsumoto, H.; Ogura, H.; Shimizu, K.; Ikeda, M.; Hirose, T.; Matsuura, H.; Kang, S.; Takahashi, K.; Tanaka, T.; Shimazu, T. The clinical importance of a cytokine network in the acute phase of sepsis. Sci. Rep. 2018, 8, 13995. [Google Scholar] [CrossRef] [PubMed]
  46. Miguel-Bayarri, V.; Casanoves-Laparra, E.; Pallás-Beneyto, L.; Sancho-Chinesta, S.; Martín-Osorio, L.; Tormo-Calandín, C.; Bautista-Rentero, D. Prognostic value of the biomarkers procalcitonin, interleukin-6 and C-reactive protein in severe sepsis. Med. Intensiv. 2012, 36, 556–562. [Google Scholar] [CrossRef] [PubMed]
  47. Oberholzer, A.; Souza, S.M.; Tschoeke, S.K.; Oberholzer, C.; Abouhamze, A.; Pribble, J.P.; Moldawer, L.L. Plasma cytokine measurements augment prognostic scores as indicators of outcome in patients with severe sepsis. Shock 2005, 23, 488–493. [Google Scholar]
  48. Phua, J.; Koay, E.S.C.; Lee, K.H. Lactate, procalcitonin, and amino-terminal pro-b-type natriuretic peptide versus cytokine measurements and clinical severity scores for prognostication in septic shock. Shock 2008, 29, 328–333. [Google Scholar] [CrossRef]
  49. Ricarte-Bratti, J.P.; Jaime-Albarrán, N.; Montrull, H.L.; Brizuela, N.Y. IL-6, MMP 3 and prognosis in previously healthy sepsis patients. Rev. Fac. Cien. Med. Univ. Nac. Cordoba 2017, 74, 99–106. [Google Scholar] [CrossRef]
  50. Ríos-Toro, J.-J.; Márquez-Coello, M.; García-Álvarez, J.-M.; Martín-Aspas, A.; Rivera-Fernández, R.; de Benito, A.S.; Girón-González, J.-A. Soluble membrane receptors, interleukin 6, procalcitonin and C reactive protein as prognostic markers in patients with severe sepsis and septic shock. PLoS ONE 2017, 12, e0175254. [Google Scholar] [CrossRef]
  51. Siddiqui, S.; Gurung, R.L.; Liu, S.; Seet, E.C.P.; Lim, S.C. Genetic Polymorphisms and Cytokine Profile of Different Ethnicities in Septic Shock Patients, and their Association with Mortality. Indian J. Crit. Care Med. 2019, 23, 135–138. [Google Scholar] [CrossRef] [PubMed]
  52. Song, J.; Park, D.W.; Moon, S.; Cho, H.-J.; Park, J.H.; Seok, H.; Choi, W.S. Diagnostic and prognostic value of interleukin-6, pentraxin 3, and procalcitonin levels among sepsis and septic shock patients: A prospective controlled study according to the Sepsis-3 definitions. BMC Infect. Dis. 2019, 19, 968. [Google Scholar] [CrossRef]
  53. Takahashi, W.; Nakada, T.A.; Yazaki, M.; Oda, S. Interleukin-6 Levels Act as a Diagnostic Marker for Infection and a Prognostic Marker in Patients with Organ Dysfunction in Intensive Care Units. Shock 2016, 46, 254–260. [Google Scholar] [CrossRef]
  54. Thao, P.T.N.; Tra, T.T.; Son, N.T.; Wada, K. Reduction in the IL-6 level at 24 h after admission to the intensive care unit is a survival predictor for Vietnamese patients with sepsis and septic shock: A prospective study. BMC Emerg. Med. 2018, 18, 39. [Google Scholar] [CrossRef]
  55. Turan, Y.B. The role of proadrenomedullin, interleukin 6 and CD64 in the diagnosis and prognosis of septic shock. BMC Anesthesiol. 2023, 23, 278. [Google Scholar] [CrossRef] [PubMed]
  56. Vivas, M.C.; Villamarin, C.; Guerrero, H.F.V.; Tascon, A.J.; Valderrama-Aguirre, A. Plasma interleukin-6 levels correlate with survival in patients with bacterial sepsis and septic shock. Interv. Med. Appl. Sci. 2021, 11, 224–230. [Google Scholar] [CrossRef] [PubMed]
  57. Weidhase, L.; Wellhöfer, D.; Schulze, G.; Kaiser, T.; Drogies, T.; Wurst, U.; Petros, S. Is Interleukin-6 a better predictor of successful antibiotic therapy than procalcitonin and C-reactive protein? A single center study in critically ill adults. BMC Infect. Dis. 2019, 19, 150. [Google Scholar] [CrossRef]
  58. Wu, C.; Ma, J.; Yang, H.; Zhang, J.; Sun, C.; Lei, Y.; Liu, M.; Cao, J. Interleukin-37 as a biomarker of mortality risk in patients with sepsis. J. Infect. 2021, 82, 346–354. [Google Scholar] [CrossRef] [PubMed]
  59. Wu, H.-P.; Chen, C.-K.; Chung, K.; Jiang, B.-Y.; Yu, T.-J.; Chuang, D.-Y. Plasma transforming growth factor-β1 level in patients with severe community-acquired pneumonia and association with disease severity. J. Formos. Med. Assoc. 2009, 108, 20–27. [Google Scholar] [CrossRef]
  60. Xie, Y.; Li, B.; Lin, Y.; Shi, F.; Chen, W.; Wu, W.; Zhang, W.; Fei, Y.; Zou, S.; Yao, C. Combining Blood-Based Biomarkers to Predict Mortality of Sepsis at Arrival at the Emergency Department. Med Sci. Monit. 2020, 27, e929527. [Google Scholar] [CrossRef]
  61. Xie, Y.; Zhuang, D.; Chen, H.; Zou, S.; Chen, W.; Chen, Y. 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. 2022, 42, 77–85. [Google Scholar] [CrossRef] [PubMed]
  62. Yu, B.; Chen, M.; Zhang, Y.; Cao, Y.; Yang, J.; Wei, B.; Wang, J. Diagnostic and Prognostic Value of Interleukin-6 in Emergency Department Sepsis Patients. Infect. Drug Resist. 2022, ume 15, 5557–5566. [Google Scholar] [CrossRef]
  63. Zhang, Y.; Khalid, S.; Jiang, L. Diagnostic and predictive performance of biomarkers in patients with sepsis in an intensive care unit. J. Int. Med. Res. 2018, 47, 44–58. [Google Scholar] [CrossRef] [PubMed]
  64. Zhao, Y.; Li, C.; Jia, Y. Evaluation of the Mortality in Emergency Department Sepsis score combined with procalcitonin in septic patients. Am. J. Emerg. Med. 2013, 31, 1086–1091. [Google Scholar] [CrossRef]
  65. Molano-Franco, D.; Arevalo-Rodriguez, I.; Muriel, A.; del Campo-Albendea, L.; Fernández-García, S.; Alvarez-Méndez, A.; Simancas-Racines, D.; Viteri, A.; Sanchez, G.; Fernandez-Felix, B.; et al. Basal procalcitonin, C-reactive protein, interleukin-6, and presepsin for prediction of mortality in critically ill septic patients: A systematic review and meta-analysis. Diagn. Progn. Res. 2023, 7, 15. [Google Scholar] [CrossRef] [PubMed]
  66. Li, X.-Y.; Liu, M.; Fu, Y.-J.; Jiang, Y.-J.; Zhang, Z.-N. Alterations in levels of cytokine following treatment to predict outcome of sepsis: A meta-analysis. Cytokine 2022, 161, 156056. [Google Scholar] [CrossRef]
  67. Charles, P.E.; Tinel, C.; Barbar, S.; Aho, S.; Prin, S.; Doise, J.M.; Olsson, N.O.; Blettery, B.; Quenot, J.P. Procalcitonin kinetics within the first days of sepsis: Relationship with the appropriateness of antibiotic therapy and the outcome. Crit. Care 2009, 13, R38. [Google Scholar] [CrossRef]
  68. Karlsson, S.; Heikkinen, M.; Pettilä, V.; Alila, S.; Väisänen, S.; Pulkki, K.; Kolho, E.; Ruokonen, E.; the Finnsepsis Study Group. Predictive value of procalcitonin decrease in patients with severe sepsis: A prospective observational study. Crit. Care 2010, 14, R205. [Google Scholar] [CrossRef]
  69. Aggarwal, A.; Singh, S.; Singh, R.; Poddar, B.; Baronia, A. Procalcitonin kinetics as a prognostic marker in severe sepsis/septic shock. Indian J. Crit. Care Med. 2015, 19, 140–146. [Google Scholar] [CrossRef]
  70. Schuetz, P.; Maurer, P.; Punjabi, V.; Desai, A.; Amin, D.N.; Gluck, E. Procalcitonin decrease over 72 hours in US critical care units predicts fatal outcome in sepsis patients. Crit. Care 2013, 17, R115. [Google Scholar] [CrossRef]
  71. Shi, Y.; Peng, J.-M.; Hu, X.-Y.; Wang, Y. The utility of initial procalcitonin and procalcitonin clearance for prediction of bacterial infection and outcome in critically ill patients with autoimmune diseases: A prospective observational study. BMC Anesthesiol. 2015, 15, 137. [Google Scholar] [CrossRef] [PubMed]
  72. Nor, M.B.M.; Ralib, A.M. Procalcitonin clearance for early prediction of survival in critically ill patients with severe sepsis. Crit. Care Res. Pract. 2014, 2014, 819034. [Google Scholar] [CrossRef]
  73. Pieralli, F.; Vannucchi, V.; Mancini, A.; Antonielli, E.; Luise, F.; Sammicheli, L.; Turchi, V.; Para, O.; Bacci, F.; Nozzoli, C. Procalcitonin Kinetics in the First 72 Hours Predicts 30-Day Mortality in Severely Ill Septic Patients Admitted to an Intermediate Care Unit. J. Clin. Med. Res. 2015, 7, 706–713. [Google Scholar] [CrossRef] [PubMed]
  74. Lee, M.J.; Bae, J.; Lee, J.H.; Park, Y.J.; Lee, H.A.R.; Mun, S.; Kim, Y.-S.; Yune, C.J.; Chung, T.N.; Kim, K. Serial Change of Endotoxin Tolerance in a Polymicrobial Sepsis Model. Int. J. Mol. Sci. 2022, 23, 6581. [Google Scholar] [CrossRef] [PubMed]
  75. Levy, M.M.; Fink, M.P.; Marshall, J.C.; Abraham, E.; Angus, D.; Cook, D.; Cohen, J.; Opal, S.M.; Vincent, J.-L.; Ramsay, G. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit. Care Med. 2003, 31, 1250–1256. [Google Scholar] [CrossRef]
  76. Ershler, W.B.; Keller, E.T. Age-associated increased interleukin-6 gene expression, late-life diseases, and frailty. Annu. Rev. Med. 2000, 51, 245–270. [Google Scholar] [CrossRef]
  77. Ferrucci, L.; Fabbri, E. Inflammageing: Chronic inflammation in ageing, cardiovascular disease, and frailty. Nat. Rev. Cardiol. 2018, 15, 505–522. [Google Scholar] [CrossRef]
  78. Puzianowska-Kuźnicka, M.; Owczarz, M.; Wieczorowska-Tobis, K.; Nadrowski, P.; Chudek, J.; Slusarczyk, P.; Skalska, A.; Jonas, M.; Franek, E.; Mossakowska, M. Interleukin-6 and C-reactive protein, successful aging, and mortality: The PolSenior study. Immun. Ageing 2016, 13, 21. [Google Scholar] [CrossRef]
  79. Montoya-Ruiz, C.; Jaimes, F.A.; Rugeles, M.T.; López, J.Á.; Bedoya, G.; Velilla, P.A. Variants in LTA, TNF, IL1B and IL10 genes associated with the clinical course of sepsis. Immunol. Res. 2016, 64, 1168–1178. [Google Scholar] [CrossRef]
  80. Pan, W.; Zhang, A.Q.; Yue, C.L.; Gao, J.W.; Zeng, L.; Gu, W.; Jiang, J.X. Association between interleukin-10 polymorphisms and sepsis: A meta-analysis. Epidemiol. Infect. 2015, 143, 366–375. [Google Scholar] [CrossRef]
  81. Gao, J.-W.; Zhang, A.-Q.; Pan, W.; Yue, C.-L.; Zeng, L.; Gu, W.; Jiang, J. Association between IL-6-174G/C Polymorphism and the Risk of Sepsis and Mortality: A Systematic Review and Meta-Analysis. PLoS ONE 2015, 10, e0118843. [Google Scholar] [CrossRef] [PubMed]
Biomolecules 15 00407 g001
Figure 1. Study selection process according to PRISMA guidelines.
Figure 1. Study selection process according to PRISMA guidelines.
Biomolecules 15 00407 g001
Biomolecules 15 00407 g002
Figure 2. Forest plot examining association between baseline IL-6 values and 28–30-day mortality in sepsis patients using random effects model [38,40,41,42,43,44,51,52,58,59,60,61,62,63,64].
Figure 2. Forest plot examining association between baseline IL-6 values and 28–30-day mortality in sepsis patients using random effects model [38,40,41,42,43,44,51,52,58,59,60,61,62,63,64].
Biomolecules 15 00407 g002
Biomolecules 15 00407 g003
Figure 3. Forest plot examining association between IL-6 levels within first 96 h after admission (IL-6 0–96 h) and 28–30-day mortality in sepsis patients using random effects model [34,38,61].
Figure 3. Forest plot examining association between IL-6 levels within first 96 h after admission (IL-6 0–96 h) and 28–30-day mortality in sepsis patients using random effects model [34,38,61].
Biomolecules 15 00407 g003
Biomolecules 15 00407 g004
Figure 4. Forest plot assessing discriminatory ability of baseline IL-6 plasma values for predicting 28–30-day mortality [37,43,44,46,48,53,55,58,59,60,63,64].
Figure 4. Forest plot assessing discriminatory ability of baseline IL-6 plasma values for predicting 28–30-day mortality [37,43,44,46,48,53,55,58,59,60,63,64].
Biomolecules 15 00407 g004
Biomolecules 15 00407 g005
Figure 5. Forest plot assessing discriminatory ability of IL-6 clearance at 96 h post admission [46,53].
Figure 5. Forest plot assessing discriminatory ability of IL-6 clearance at 96 h post admission [46,53].
Biomolecules 15 00407 g005
Table 1. Overview of included studies; ICU = intensive care unit; S = survivors; NS = non-survivors; SD = standard deviation; NR = not reported; DTACM = Department of Traumatology and Acute Critical Medicine.
Table 1. Overview of included studies; ICU = intensive care unit; S = survivors; NS = non-survivors; SD = standard deviation; NR = not reported; DTACM = Department of Traumatology and Acute Critical Medicine.
Study IDCountryClinical SettingPatient
Recruitment		NEvents/NAge (years)Sex (% of Males)OutcomeSepsis
Criteria		Sepsis Origin
Andaluz-Ojeda 2012 [34]SpainICUProspective2912/2966.1; mean58.70%28–30-day mortalityLevy 2001Mixed
Belli 2022 [35]ItalyICUProspective3515/3559 (48–60); median (25th–75th)60%28–30-day mortalitySinger 2015Mixed
Beneyto 2016 [36]SpainICUProspective20352/20365; mean64%In-hospital mortalityLevy 2001Mixed
Eidt 2016 [37]BrazilICUProspective4821/4861.4 (18.7) in severe sepsis; 65.9 (17.8)
in septic shock; mean (SD)		50%ICU mortalityLevy 2001NR
Frencken 2017 [38]NetherlandsICUProspective708226/70863 (53–72); median (25th–75th)60%4-day, 28-day, and 1-year mortalitySinger 2015Mixed
Jekarl 2013 [39]South KoreaMixedProspective7815/7862.1 (19.9); mean (SD)45%28–30-day mortalityLevy 2001Mixed
Jiang 2019 [40]ChinaICUProspective19888/19869.9 (28–91) in S vs. 68.7 (18;91) in NS; median (25th–75th)81.80%28–30-day mortalitySinger 2015Mixed
Karamouzos 2021 [41]GreeceICURetrospective12846/12872.4 (15) in S vs. 77.3 (10) in NS;
mean (SD)		43.80%28–30-day mortalitySinger 2015Mixed
Karampela 2022 [42]GreeceICUProspective10230/10264.7 (15.6); mean (SD)55.90%28–30-day mortalitySinger 2015Mixed
Liu S. 2021 [43]ChinaICURetrospective26478/264 52.9 (12.6); mean (SD)64%28–30-day mortalitySinger 2015Mixed
Liu J. 2021 [44]ChinaICUProspective6617/6669 (54–82) in S vs. 77 (63–84) in NS;
median (IQR)		59%28–30-day mortalitySinger 2015Mixed
Matsumoto 2018 [45]JapanDTACMRetrospective317/3173.0 (65.0–81.0); median (IQR)74%28–30-day mortalitySinger 2015Mixed
Miguel-Bayarri 2012 [46]SpainICUProspective8127/8162, median; no other details54%28–30-day mortalityLevy 2001Mixed
Oberholzer 2005 [47]USAMixedProspective12439/12458.3 (17.5); mean (SD)55%28–30-day mortalityOtherNR
Phua 2008 [48]SingaporeICUProspective7230/7255 (16) in S vs. 54 (17) in NS; mean (SD)64%28–30-day mortalityLevy 2001Mixed
Ricarte-Bratti 2017 [49]ArgentinaICUProspective4821/4853.8 (17.1) in S vs. 73.2 (8.9) in NS;
mean (SD)		46%28–30-day mortalitySinger 2015Mixed
Rios-Toro 2017 [50]SpainICUProspective5021/5068 (53–75); median (25th–75th)72%28–30-day mortalityLevy 2001Mixed
Siddiqui 2019 [51]SingaporeICUProspective198NR63.0 (49.5;73.5); 54.0 (45.0;64.0); 61.0 (40.0;68.8); mean (SD)60.10%28–30-day mortalityOtherNR
Song 2019 [52]South KoreaERProspective97NR75 (42;98); median (25th–75th)56.00%28–30-day mortalitySinger 2015Mixed
Takahashi 2016 [53]JapanICUProspective8515/8568 (59–77); median (25th–75th)62.40%28–30-day mortalitySinger 2015Mixed
Thao 2018 [54]VietnamICUProspective12375/12362 (46–75) in S; 54 (43–73) in NS;
median (25th–75th)		NRIn-hospital mortalityLevy 2001 Mixed
Turan 2023 [55]TurkeyICUProspective6039/6060 (56–68) in S vs. 74 (62–80) in NS;
median (IQR)		60.00%28–30-day mortalitySinger 2015Mixed
Vivas 2021 [56]ColombiaICUProspective6210/6253 (19.47); mean (SD)59.67%In-hospital mortalitySinger 2015Mixed
Weidhase 2019 [57]GermanyICURetrospective328118/32864 [54;73] in S vs. 62 [55;69] in NS;
median (IQR)		63.10%In-hospital mortalityLevy 2001Mixed
Wu C.X. 2021 [58]ChinaICUProspective11451/11471 (60;81); median (25th–75th)72.90%28–30-day mortalitySinger 2015Mixed
Wu H.P. 2009 [59]ChinaMixedProspective6314/63 (22.2%)70.0 (2.0) in S vs. 69.1 (4.1) in NS;
mean (SD)		63.50%28–30-day mortalityOtherRespiratory
Xie 2021 [60]ChinaERRetrospective9023/9072 (26–97) in S vs. 77 (50–97) in NS;
median (25th–75th)		64%28–30-day mortalitySinger 2015Mixed
Xie 2023 [61]ChinaERRetrospective36753/36771 (19–98) in S vs. 80 (46–97) in NS;
median (25th–75th)		65.90%28–30-day mortalitySinger 2015Mixed
Yu 2022 [62]ChinaERProspective63NR79 (34–95); median (25th–75th)63.50%28–30-day mortalitySinger 2015NR
Zhang 2019 [63]ChinaICURetrospective15016/66 in sepsis group and 48/84 in septic shock group70 (24–91) in sepsis group vs. 74.5 (24–89) in septic shock group; median (25th–75th)63.40%28–30-day mortalitySinger 2015Mixed
Zhao 2013 [64]ChinaERProspective501134/50473 (58–79) in S vs. 77 (65–83) in NS;
median (IQR)		55.70%28–30-day mortalityLevy 2001Mixed
Table 2. IL-6 values and mortality: effect measures. HR = hazard ratio; OR = odds ratio; D = day; NR = not reported.
Table 2. IL-6 values and mortality: effect measures. HR = hazard ratio; OR = odds ratio; D = day; NR = not reported.
Study IDNEvents/NIL-6 Values and Mortality: Effect Measures (95% CI)Model Adjusted by
Andaluz-Ojeda 2012 [34]2912Adjusted HR at D3: 1.86 (1.08–3.20); D28: 2.00 (1.22−3.27)APACHE II, other biomarkers
Belli 2022 [35]3515HR = 1.000 (1.000–1.000)Univariable
Beneyto 2016 [36]20352Admission: log IL-6 OR = 1.62 (1.24–2.13);
D3: log IL-6 OR = 2.69 (1.64–4.40)		Age, sex, APACHE II, SOFA, other biomarkers
Eidt 2016 [37]4821NRAge, sex, lactate, other biomarkers
Frencken 2017 [38]708226Admission: RR = 1.13 (0.91–1.41);
D4: RR = 1.03 (0.86–1.23)		Age, Charlson comorbidity index, immunodeficiency, site of infection
Jekarl 2013 [39]7815NROther biomarkers
Jiang 2019 [40]19888OR = 1.033 (1.007–1.061)SOFA, other biomarkers such as EPO, hepcidin, ferritin, sTfR/log ferritin
Karamouzos 2021 [41]12846OR = 1.002 (0.996–1.008)Microorganism, MDR status, type of infection, cytokines
Karampela 2022 [42]10230HR = 1.70 (1.05–2.74)Age, Sex, BMI, APACHE-II, presence of septic shock
Liu S. 2021 [43]26478OR = 1.017 (1.005–1.028)Age, sex, BMI, SBP, APACHE-II, SOFA
Liu J. 2021 [44]6617OR = 1.001 (1.000–1.001)Age, SOFA, other biomarkers
Matsumoto 2018 [45]317Maximum values from three days (D1, D2, D4): 19.62 (3.47–110.80)SOFA
Miguel-Bayarri 2012 [46]8127Admission: log IL-6 OR = 1.98 (1.27–3.09); D3: log IL-6 OR = 2.6 (1.43–4.71)Age, sex, development of MOF, other biomarkers, SOFA, APACHE-II
Oberholzer 2005 [47]12439Not significant (p-value)APACHE-II, age, treatments, baseline MOD
Phua 2008 [48]7230NRAPACHE-II
Ricarte-Bratti 2017 [49]4821NRUnivariable
Rios-Toro 2017 [50]5021NRAge, SOFA, APACHE II, other biomarkers
Siddiqui 2019 [51]198NRHR = 1.46 (1.11−1.92)Age, sex, surgical methods, qSofa
Song 2019 [52]97NRHR = 1.001 (1.000–1.002)APACHE-II, SOFA, pentraxin, lactate
Takahashi 2016 [53]8515NRAge, sex, SOFA
Thao 2018 [54]12375IL-6 clearance in 24 h: ≥86%, OR = 5.67 (1.27–25.3); Il-6 clearance between 85% and 50%, OR = 1.86 (0.44–7.94)Age, sex, BUN, Creatinine, aPTT, pH
Turan 2023 [55]6039NRAge, sex, diagnosis, SOFA, APACHE II, other biomarkers
Vivas 2021 [56]6210NRAge, site of infection
Weidhase 2019 [57]328118NRNR
Wu C.X. 2021 [58]11451OR = 1.66 (0.67−4.10)SOFA and IL-37
Wu H.P. 2009 [59]6314OR = 1.00 (0.99−1.01)APACHE-II, septic shock, gastrointestinal bleeding, IL-10 and TGF-β1
Xie 2021 [60]9023OR = 1.000 (1.000–1.001)Lactate, neutrophil-to-WBC ratio
Table 3. Study-specific and pooled effect estimates of association between baseline IL-6 levels and mortality; CI = confidence interval.
Table 3. Study-specific and pooled effect estimates of association between baseline IL-6 levels and mortality; CI = confidence interval.
StudyEstimateStandard ErrorLower Limit of 95% CIUpper Limit of 95% CIzp-ValueWeight (%)—FixedWeight (%)—Random
Frencken 2017 [38]1.130.1270.8811.379 0.0010.004
Jiang 2019 [40]1.0330.0141.0061.060 0.0820.33
Karamouzos 2021 [41]1.0020.0030.9961.008 1.795.32
Karampela 2022 [42]1.70.4270.8632.537 0.0090.004
Liu S. 2021 [43]1.0170.0061.0051.029 0.451.68
Liu J. 2021 [44]1.0010.0010.9991.003 16.1414.95
Siddiqui 2019 [51]1.460.2021.0641.856 0.0010.001
Song 2019 [52]1.0010.0010.9991.003 16.1414.95
Wu C.X. 2021 [58]1.660.8280.03713.283 0.0010.001
Wu H.P. 2009 [59]10.0050.991.01 0.652.33
Xie 2021 [60]10.0010.9981.002 16.1414.95
Xie 2023 [61]10.0010.9981.002 16.1414.95
Yu 2022 [62]0.9990.0010.9971.001 16.1414.95
Zhang 2019 [63]1.020.0111.04 0.160.64
Zhao 2013 [64]1.0020.00111.004 16.1414.95
Total
(fixed effects)		1.001011.0012490.434<0.001100100
Total
(random effects)		1.0010.0010.9991.0031230.89<0.001100100
Table 4. Study-specific and pooled effect estimates of association between IL-6 clearance and mortality; CI = confidence interval.
Table 4. Study-specific and pooled effect estimates of association between IL-6 clearance and mortality; CI = confidence interval.
StudyEstimateStandard ErrorLower Limit of 95% CIUpper Limit of 95% CIzp-ValueWeight (%)—FixedWeight (%)—Random
Andaluz-Ojeda 2012 [34]1.860.5250.8312.889 0.0010.82
Frencken 2017 [38]1.030.0920.851.21 0.04720.12
Xie 2023 [61]1.0070.0021.0031.011 99.9579.06
Total
(fixed effects)		1.0070.0021.0031.011503.634<0.001100100
Total
(random effects)		1.0190.0480.9251.11221.316<0.001100100
Table 5. Discriminatory ability of baseline IL-6 values, according to area under receiver operating characteristics.
Table 5. Discriminatory ability of baseline IL-6 values, according to area under receiver operating characteristics.
StudyROC AreaStandard ErrorLower Limit of 95% CIUpper Limit of 95% CIZp-ValueWeight (%)—FixedWeight (%)—Random
Eidt 2016 [37]0.6690.01150.6460.692 0.548.22
Liu S. 2021 [43]0.8490.00140.8460.852 36.728.42
Liu J. 2021 [44]0.7850.00860.7680.802 0.978.31
Miguel-Bayarri 2012 [46]0.740.00650.7270.753 1.708.36
Phua 2008 [48]0.770.00720.7560.784 1.398.34
Takahashi 2016 [53]0.6540.00780.6390.669 1.188.33
Turan 2023 [55]0.5730.00980.5540.592 0.758.28
Wu C.X. 2021 [58]0.6160.0050.6060.626 2.888.39
Wu H.P. 2009 [59]0.7140.01130.6920.736 0.568.23
Xie 2021 [60]0.6750.00760.660.69 1.258.33
Zhang 2019 [63]0.6750.00590.6630.687 2.078.37
Zhao 2013 [64]0.6920.00120.690.694 49.988.42
Total
(fixed effects)		0.7480.00080.7470.75882.121<0.001100100
Total
(random effects)		0.7010.02110.660.74233.266<0.001100100
Table 6. Discriminatory ability of IL-6 clearance according to area under receiver operating Characteristics.
Table 6. Discriminatory ability of IL-6 clearance according to area under receiver operating Characteristics.
StudyROC AreaStandard ErrorLower Limit of 95% CIUpper Limit of 95% CIZp-ValueWeight (%)—FixedWeight (%)—Random
Miguel-Bayarri 2012 [46]0.860.040.7820.938 76.4267.92
Takahashi 2016 [53]0.760.0720.6190.901 23.5832.08
Total (fixed effects)0.8360.0350.7680.90523.921<0.001100100
Total (random effects)0.8280.04670.7360.91917.737<0.001100100
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Varga, N.-I.; Bagiu, I.C.; Vulcanescu, D.D.; Lazureanu, V.; Turaiche, M.; Rosca, O.; Bota, A.V.; Horhat, F.G. IL-6 Baseline Values and Dynamic Changes in Predicting Sepsis Mortality: A Systematic Review and Meta-Analysis. Biomolecules 2025, 15, 407. https://doi.org/10.3390/biom15030407

AMA Style

Varga N-I, Bagiu IC, Vulcanescu DD, Lazureanu V, Turaiche M, Rosca O, Bota AV, Horhat FG. IL-6 Baseline Values and Dynamic Changes in Predicting Sepsis Mortality: A Systematic Review and Meta-Analysis. Biomolecules. 2025; 15(3):407. https://doi.org/10.3390/biom15030407

Chicago/Turabian Style

Varga, Norberth-Istvan, Iulia Cristina Bagiu, Dan Dumitru Vulcanescu, Voichita Lazureanu, Mirela Turaiche, Ovidiu Rosca, Adrian Vasile Bota, and Florin George Horhat. 2025. "IL-6 Baseline Values and Dynamic Changes in Predicting Sepsis Mortality: A Systematic Review and Meta-Analysis" Biomolecules 15, no. 3: 407. https://doi.org/10.3390/biom15030407

APA Style

Varga, N.-I., Bagiu, I. C., Vulcanescu, D. D., Lazureanu, V., Turaiche, M., Rosca, O., Bota, A. V., & Horhat, F. G. (2025). IL-6 Baseline Values and Dynamic Changes in Predicting Sepsis Mortality: A Systematic Review and Meta-Analysis. Biomolecules, 15(3), 407. https://doi.org/10.3390/biom15030407

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop