A comprehensive search approach was crafted to identify relevant articles across major databases, specifically Web of Science, PubMed, Scopus, Web of Science, and the Cochrane Library, from January 2020 to September 2024. Appendix 1 provides a detailed description of this strategy. Additionally, Google Scholar was reviewed to ensure any additional pertinent studies were captured. The search terms were thoughtfully identified and crafted using a composite of the following keywords: “COVID-19,” “S100 Proteins,” “Pneumonia,” and “Interleukins.” The methodology for this systematic review adhered to the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)” guidelines (as illustrated in Figure 1).

The review included studies investigating the expression of biomarkers, specifically S100 proteins and interleukin markers, in COVID-19 patients, with and without pneumonia or organ failure. The selected articles comprised a range of study designs, including interventional trials and observational studies, such as prospective and retrospective cohort studies, nested studies, case-control studies (both matched and unmatched), cross-sectional analyses, and case reports. Exclusion criteria encompassed publications of non-pertinent formats (e.g., letters to the editor, editorials, or comments), non-English articles, and studies whose data did not align with the primary focus of the review.

The study screening process was completed in two phases. Firstly, two reviewers (VD and WAWS) independently reviewed the titles and abstracts of the retrieved studies to identify those meeting the inclusion criteria. In the second phase, the same reviewers examined the full texts for potentially relevant studies, including only those that aligned with the criteria. Any disagreements between researchers were resolved through discussion and consensus, with a third reviewer (MNS) stepping in for arbitration. For this meta-analysis, disease severity was operationalised according to the categories reported in the included studies. Specifically, we compared biomarker profiles between (i) COVID-19 patients with pneumonia and (ii) COVID-19 patients without pneumonia. In addition, where studies reported data, we included analyses of (iii) COVID-19 patients with organ failure versus those without organ failure. Several studies also provided data on (iv) healthy individuals, who were included as additional comparator groups to provide context for interpreting biomarker elevations.

An MS Excel-based data extraction form was developed following detailed discussions with the review team to ensure consistency. The form captured comprehensive information, organized into key categories: study identifiers (“author, study year, publication year”), article origin details (“country, study setting, publication type”), and study specifics (“design, patient demographics, baseline characteristics, and main findings”). Two researchers independently conducted data extraction, and any discrepancies or inconsistencies were resolved through discussion, consensus-building, or by involving a third team member.

The Newcastle-Ottawa Scale (NOS), a standard tool for observational studies, was used to assess the quality of the included studies. The NOS was applied separately to cohort and case-control studies, with an adapted version used for cross-sectional studies. This scale assigns up to 9 points across three domains: “selection,” “comparability,” and “outcome or exposure assessment.” Studies were classified based on their scores: low quality (0–3 points), moderate quality (4–6 points), and high quality (7–9 points).

A meta-analysis was conducted to estimate the prevalence of COVID-19, with or without associated pneumonia or organ failure, in relation to S100 proteins and interleukin markers. Forest plots were generated to visually represent the collated data and the effect estimates across studies. Heterogeneity was evaluated using the χ² test with a significance level of 5%, and effect sizes were expressed as standardized mean differences (SMDs). The degree of heterogeneity was quantified using the I² statistic, with values above 50% signifying moderate heterogeneity among the included studies. All analyses were performed using RevMan software. Patients were stratified by disease severity into subgroups, specifically COVID-19 with or without pneumonia or organ failure, while several studies also included healthy controls as comparator groups. This classification enabled consistent subgroup analyses; however, it should be acknowledged that the included studies did not uniformly apply standardized severity definitions, such as those proposed by the World Health Organization (WHO) or the National Institutes of Health (NIH).

Each figure represents an independent subgroup comparison—pneumonia versus healthy controls, and pneumonia versus non-pneumonia COVID-19 cases, respectively. These analyses are presented separately and are not intended for direct cross-comparison.

The NOS was adapted for cohort, case-control, and cross-sectional studies to evaluate study quality (Figures 6A–C). Out of 51 studies, 14 (27%) had low risk, 35 (69%) moderate risk, and 2 (4%) high risk of bias, with quality scores ranging from 3 to 8 points.

A major observation in this meta-analysis is the significant relationship between elevated interleukin levels, mainly IL-6 and IL-10, and adverse COVID-19 outcomes. IL-6 is a well-established pro-inflammatory cytokine that contributes to the inflammatory cascade, a hallmark of acute COVID-19 patients, particularly in patients requiring advanced respiratory support or critical care. Consistent with earlier studies, such as the RECOVERY trial [62, 63], our results demonstrate that IL-6 levels are significantly higher in COVID-19 patients with pneumonia compared with those without pneumonia, and also compared with healthy controls, confirming its role in exacerbating hyperinflammatory responses.

Despite its anti-inflammatory nature, IL-10 is also markedly increased in acute COVID-19 patients, reflecting a broader dysregulation of the immune system [64]. Our findings revealed a notable elevation of IL-10 in COVID-19 patients who developed organ failure compared with those without organ failure, consistent with previous literature that highlights IL-10’s role in severe disease progression [11]. While IL-6 is widely recognized as a driver of pro-inflammatory activity, the rise in IL-10 levels could signify the body’s attempt to regulate the excessive inflammation; although, it appears insufficient to counterbalance the immune dysfunction. The observed heterogeneity in IL-10 levels across different studies highlights the complexity of immune responses in COVID-19, indicating that IL-10 may be more indicative of generalized immune dysregulation than inflammation alone. This aligns with other studies indicating that immune cell subsets, particularly CD8⁺ T cells, play a crucial role in COVID-19 outcomes, with altered lymphocyte counts correlating to worse prognosis [22].

This meta-analysis also highlights the involvement of S100 proteins, particularly S100B and calprotectin (S100A8/A9), in COVID-19 pathophysiology. Both markers are well known for their roles in modulating immune and inflammatory responses, and their elevated values in COVID-19 cases were consistently observed in patients with more severe disease presentations, supporting their potential role as indicators of severity [65]. S100B, a protein typically linked to neuroinflammation, was found to be significantly elevated in COVID-19 pneumonia patients compared with healthy controls (SMD = 0.51, 95% CI 0.19–0.83, p = 0.002). When comparing COVID-19 patients with and without pneumonia, S100B was also higher in the pneumonia group (SMD = 0.30, 95% CI 0.02–0.57, p = 0.04). These findings suggest that S100B is not only elevated in COVID-19 but may also reflect greater inflammatory burden in patients who progress to pneumonia [56]. Although the included studies largely reported biomarkers independently, the strength of this meta-analysis lies in synthesizing IL-6, IL-10, IL-8, and S100B together. By evaluating these markers side by side, our findings provide a broader multi-marker perspective that reflects different aspects of the inflammatory and tissue injury response in COVID-19, and offers a basis for future studies to assess their combined predictive value.

Calprotectin (S100A8/A9), recognized for its role in neutrophil activation and systemic inflammation, also showed elevated levels in patients requiring intensive care. This finding corroborates earlier studies, which have demonstrated that higher calprotectin levels are predictive of poor outcomes in respiratory infections [55]. The elevated levels of S100A8/A9 in COVID-19 patients, particularly those with severe pneumonia and multi-organ failure, suggest its potential as a biomarker for disease progression, helping to predict which patients may require intensive care or more aggressive treatment [66]. This aligns with studies that demonstrate the relevance of these biomarkers in predicting complications like thromboinflammation and respiratory failure [25, 41].

The variability in findings related to IL-6 and IL-10, particularly in patients with organ failure, indicates differences in cytokine expression in relation to clinical severity. Such variability is understandable, given the diversity of patient populations, underlying health conditions, and treatment regimens included in the analysis. Nonetheless, it underscores the need for more consistent methodologies for assessing cytokine levels in clinical practice. For example, the substantial heterogeneity (I² = 90%) seen in IL-10 studies may reflect differences in patient demographics and the timing of cytokine measurement, which could contribute to the differing outcomes observed [6, 11].

In contrast, the low heterogeneity (I² < 30%) across studies examining IL-6 and S100B levels in patients with pneumonia indicates more uniform findings [67, 68]. This consistency strengthens the case for using these markers as reliable predictors of disease severity, particularly in patients presenting with COVID-19 pneumonia. The low variability suggests that these biomarkers may have greater utility in clinical settings for the early identification of high-risk patients. Additionally, biomarkers such as CCL17, CRP, and procalcitonin, along with the adiponectin-to-leptin ratio, have shown promise in distinguishing between mild and severe disease, further supporting their use in clinical practice [28, 46]. To avoid misinterpretation, it should be emphasized that analyses comparing COVID-19 pneumonia with healthy controls and those comparing pneumonia with non-pneumonia COVID-19 reflect different cohorts and therefore cannot be directly compared.

The findings of this meta-analysis have important implications for clinical practice. The consistent association of elevated IL-6, IL-10, and S100B levels with severe disease outcomes points to their potential use in stratifying COVID-19 patients based on risk [69]. Early identification of patients with heightened levels of these biomarkers could facilitate timely intervention with anti-inflammatory treatments or more intensive monitoring. For example, elevated IL-6 has been incorporated into risk stratification protocols to identify patients who may benefit from early administration of immunomodulatory therapies such as tocilizumab, with several studies reporting reduced progression to respiratory failure and improved survival in high-risk patients [62]. Elevated IL-10 has been consistently associated with adverse outcomes and is increasingly regarded as a marker for intensified monitoring and timely supportive care [70]. In contrast, although S100B shows a strong association with pneumonia and disease progression, its role in guiding early intervention strategies remains insufficiently defined [56]. Collectively, these findings indicate that while these biomarkers hold promise for early risk stratification, further prospective studies are needed to validate their utility in improving patient outcomes.

However, further research is needed to standardize the measurement of these biomarkers across different clinical settings. More rigorous and consistent protocols will enable better comparability of future studies and enhance the clinical applicability of these biomarkers. Additionally, longitudinal studies assessing the dynamics of these biomarkers over time will provide insights into their role in predicting long-term outcomes, such as the post-acute sequelae of COVID-19. As research advances, combining multi-omics approaches, such as proteomics, genomics, and metabolomics, with biomarker research could reveal new molecular pathways and therapeutic targets for managing severe COVID-19. Furthermore, comparative analyses between COVID-19 biomarkers and those of other respiratory infections could help identify specific disease markers and facilitate the development of targeted therapeutic strategies.

The strengths of this study encompass a thorough analysis that consolidates data from numerous research efforts, offering a wide-ranging assessment of S100 proteins and interleukins (IL-6, IL-10) across different stages of COVID-19 severity, thereby delivering important insights into their clinical significance. Moreover, the study reveals consistent results for crucial biomarkers, such as IL-6 and S100B, exhibiting low variability (I² < 30%) among the studies, which suggests dependable associations with COVID-19 severity, particularly in pneumonia cases. On the downside, the study also faces limitations. Incomplete reporting of demographic variables such as age, sex distribution, and BMI across the included studies limited our ability to assess their influence on biomarker levels. The examination of IL-10 levels demonstrated significant variability (I² = 90%), likely attributed to differences in study methodologies, patient demographics, or the timing of cytokine assessments, which impacts the reliability of these findings. Additionally, certain analyses, especially those regarding S100 proteins, relied on a limited number of studies, potentially diminishing the generalizability of the results and highlighting the need for further investigation.

Future research should focus on longitudinal studies that monitor S100 proteins and interleukins to gain deeper insights into their roles in COVID-19 progression and recovery. Including more diverse populations in these studies would help identify biomarker variability, enhancing the relevance of findings across different demographic and ethnic groups. Investigating the potential of IL-6, IL-10, and S100 proteins as therapeutic targets may pave the way for personalized treatments aimed at mitigating inflammation and reducing severe outcomes. Combining multi-omics approaches, such as proteomics, genomics, and metabolomics, with biomarker research could reveal new molecular pathways and targets in severe COVID-19 cases. Additionally, comparative analyses between COVID-19 biomarkers and those of other respiratory infections could help identify specific disease markers and facilitate the development of targeted therapeutic strategies.

COVID-19 severity ranges from mild infection to pneumonia, Acute Respiratory Distress Syndrome (ARDS), and organ failure (=111).

IL-6 is significantly elevated in COVID-19 patients with pneumonia versus those without (=94).

This work represents an advance in biomedical science because it consolidates evidence that IL-6, IL-10, and S100B are reliable biomarkers for assessing COVID-19 pneumonia severity and progression (=194).