The relationship between C-reactive protein and levels of various cytokines in patients with COVID-19: A systematic review and correlation analysis.

Introduction: C-reactive protein (CRP) and cytokines levels could alter in patients with coronavirus disease (COVID-19) due to the inflammatory response caused by the virus. This analysis aimed to assess the relationship between the CRP levels and the levels of various cytokines in COVID-19 patients. Materials and
Methods: We searched the databases of PubMed, Cochrane, and Web of Science for relevant articles on May 29th, 2021. Applying the inclusion/exclusion criteria, the retrieved records underwent two-phase screenings; first, a title/abstract screening process, and then, a full-text screening to find the eligible studies. Data for study variables were extracted, including the CRP levels and the levels of all reported cytokines. A strong and significant relationship between Interleukins and CRP was defined as: p ≤ 0.05, 0.7 ≤ r ≤ 1.
Results: In this study, 103 studies were included for systematic review and correlation analysis. The aggregate mean and SD of study variables were calculated and reported. The correlation between Interleukins and CRP was measured using correlation coefficient (r). It appeared that interleukin (IL)-10 has a moderate and significant relationship with CRP (p ≤ 0.05, r = 0.472). IL-10 predicted almost 10% of CRP changes.
Conclusion: This correlation analysis suggests IL-10 is moderately correlated with CRP levels in patients with COVID-19 infection. A better understanding of the pro-inflammatory markers could contribute to the implementation of therapeutic and preventive approaches. More prospective studies are suggested to explore the relationship between CRP and cytokines as potential markers for the early identification of COVID-19 progression and severity.

INTRODUCTION

The coronavirus disease (COVID‐19), caused by acute respiratory dysfunction coronavirus 2 (SARS‐CoV‐2), has been pandemic across the countries. It is still pandemic around the world. Many people experience mild to moderate symptoms, but some also suffer from severe symptoms and even death.

Common symptoms of COVID‐19 include fever, cough, myalgia, fatigue, or respiratory distress. In severe cases of the disease, dyspnea and respiratory problems occur and cause acute respiratory distress syndrome (ARDS) or multiple organs. Following the proliferation of the virus, the host's immune system is also activated to fight the virus and cure the patient, however some patients develop severe and disease such as MODS, The cause is still unknown. , Recent studies have showed that, in addition to common symptoms, untreated inflammation contributes to disease severity in COVID‐19. For those infected by SARS‐CoV‐2, some of the patients did not show dispnea or respiratory distress during the COVID‐19, indicating a multifaceted disease of COVID‐19 infection. The prognosis for patients is not fully understood, and the type and severity of the consequences of COVID‐19 may depend, inter alia, on the severity of the course of the infection, age, and comorbidities. Extreme COVID‐19 is characterized by interleukin waves, organ dysfunction disease, and destruction of many metabolic processes, including coagulation factors and blood clotting. Therefore, a reliable and appropriate test is needed to predict the severity of COVID‐19 disease. Recently, several studies have reported that C‐reactive protein (CRP) is directly associated with the severity of infection, and patients with higher CRP in the early stages of the disease are at greater risk for severe disease.

CRP, first described by Tillet and Francis, is released from the liver in response to interleukin‐6 (IL‐6) and it is an inflammatory test that is available everywhere. Many studies have also shown that the level of this biomarker is associated with the severity of influenza, and recently some studies have reported an association between CRP levels with the severity of COVID‐19 disease. Consistent with this hypothesis; high levels of inflammatory markers such as CRP, ferritin, and d‐dimer; high neutrophil‐to‐lymphocyte ratio; and high levels of cytokines and chemokines have been observed in patients with severe COVID‐19 diseases. Pathogenic inflammation, also referred to as cytokine storm, shares similarities with what was previously seen in patients infected with other types of coronaviruses, such as SARS‐CoV and Middle East respiratory syndrome coronavirus. Similar other types of COVs disease, there are more evidence that cytokine storms due to overproduction of inflammatory factors may play a role in COVID‐19 symptoms, which may be a key factor in the rapid deterioration of the disease.

A recent study showed that about 7.7% of patients with mild symptoms of COVID‐19 progress to severe disease and have a high CRP level content compared to mild disease. Clinical aspects can be interpreted more clearly when examined with biological markers such as CRP. As a result, investigating the CRP level may be critical for early identification and adequate therapy of COVID‐19‐related problems. Since the outbreak of COVID‐19, many studies have studied changes in biomarkers in the diagnosis of COVID‐19, but associations of cytokines and CRP with severity of disease have received less attention. Therefore, this systematic review and correlation analysis aimed to primarily assess the association between the CRP levels and the levels of various cytokines in COVID‐19 patients to inform future practice and clinical guidelines.

MATERIALS AND METHODS

This analysis orchestrated to reflect on the potential correlations between CRP and various cytokines levels in patients with COVID‐19 based on current evidence. To ensure the validity and authenticity of this report, we adhered to the preferred reporting items for systematic reviews and meta‐analyses (PRISMA) checklist as illustrated in Figure 1. To refrain the included studies from bias risk we utilized the Newcastle‐Ottawa Scale (NOS) with three categories (maximum scoring for each category in the following parentheses) of selection, comparability, and exposure/outcome and total scoring of 9 (see Table 1).

Figure 1

Preferred reporting items for systematic reviews and meta‐analyses 2020, flow diagram for systematic literature review

Table 1

Newcastle‐Ottawa Scale (NOS) tool for bias risk assessment of the study

1st author	Selection (out of 4)	Comparability (out of 2)	Exposure/Outcome (out of 3)	Total (out of 9)
Kozlovskaya LI	3	1	2	6
Del Valle DM	3	2	2	7
Chen W	2	2	3	7
Smilowitz NR	3	2	3	8
Gao YM	2	2	3	7
Ali N	4	2	3	9
Antwi‐Amoabeng D	2	1	2	5
Berenguer J	3	1	2	6
Bommenahalli Gowda S	3	1	2	6
Akbari H	3	2	2	7
Agostinis P	3	1	3	7
Cui N	3	2	2	7
Zhou S	3	2	2	7
Wang F	2	2	1	5
Venktaraman A	3	1	3	7
Bao J	4	1	3	8
Xu X	4	1	2	7
Alam JM	2	1	2	5
Bozkurt FT	3	1	2	6
Avila‐Nava A	3	1	2	6
Birben B	3	1	2	6
Da B	3	1	3	7
Webb BJ	4	1	1	6
Ullah W	2	2	3	7
Bawiskar N	2	2	3	7
Burian E	3	1	1	5
Cabrero‐Hernandez M	3	1	1	5
Chen X	3	1	1	5
Zhang ZL	3	1	2	6
Zeng Z	2	1	3	6
Xia G	4	1	1	6
Broman N	4	2	2	8
Carlino MV	3	2	2	7
Chaudhary R	3	2	2	7
Chen J	2	2	3	7
Chen Z	3	1	2	6
Cordova LDS	2	2	3	7
Danwang C	2	2	3	7
Zhu Z	3	2	2	7
Zhu W	3	2	2	7
Zhou YZ	3	1	3	7
Yang PH	3	1	3	7
Wu Y	3	1	3	7
Wang M	3	2	3	8
Wang JH	2	2	3	7
Vultaggio A	3	2	2	7
Tjendra Y	3	2	3	8
Tian R	3	2	3	8
Terpos E	3	1	2	6
Bagheri‐Hosseinabadi Z	3	1	3	7
Chen Q	2	2	3	7
Ji P	3	2	2	7
Khodeir MM	3	2	3	8

Sample and data

Applying a systematic approach and browsing the relevant keywords in the online databases of PubMed, Web of Science, and Cochrane, we mined all the relevant publications in English from December 2019 to May 29, 2021. We constructed our search strategy, by combining multiple keywords derived from the medical subject headings (MeSH) in PubMed and previous studies. Supplementary material 1 illustrates the search terms for each database. The search queries for PubMed are shown below:

“COVID‐19” OR “SARS‐CoV‐2” OR “SARS‐CoV2” OR “2019‐nCoV” OR “Novel Coronavirus” [title/abstract]

“CRP” OR “CRP” OR “Cytokine” OR “Interleukin” OR “Chemokine” OR “Lymphokine” OR “Interferon” OR “IFN” OR “Tumor necrosis factor” OR “TNF” OR “Monokine” [title/abstract]

“[A]” AND “[B]”

We conducted a two‐phase screening process. First, the title and abstract of the studies were evaluated according to the inclusion criteria. Then, the full‐text of the studies was carefully investigated and the eligible studies were included for the qualitative synthesis. Original studies that assessed CRP and the level of at least one cytokine in patients with COVID‐19 were included in this analysis. The exclusion criteria were as follows:

Nonoriginal articles; for example, reviews, systematic reviews, or meta‐analyses

Nonhuman studies; for example, in vitro experiments, animal trials

Literature lacking full texts or available solely in the form of abstracts or conference abstract

Case reports or case series with less than 10 participants

The experiment groups of the clinical trials in patients with COVID‐19 that utilized immunomodulatory and anti‐inflammatory treatments, such as corticosteroids and anticytokine drugs, were not included in the analysis. However, the data of the control groups of these studies were still used.

Measures and variabales

Following summarizing the included papers, cumulative distribution of participants’ characteristics including gender, mean age, severity of symptoms, advanced procedures applications, patient outcome, administered medications, comorbidities, various laboratory values, CRP level, and all the reported cytokines were collected and organized in a specifically designed sheet (Table 2). Four investigators independently prepared and cross‐checked this information. Other researchers went through the selected studies and accumulated data to avoid any probable duplications and crossovers.

Table 2

Cumulative distribution of participants'characteristics in included studies

Characteristics	Combined mean (SD)	Number of studies
Male (%)	56.02 (18.05)	103
Age (years)	56.16 (13.93)	103
Mild (%)	42.04 (36.15)	55
Severe (%)	47.82 (37.13)	79
ICU admission (%)	35.34 (34.74)	36
Oxygen supplementation (%)	50.75 (34.74)	34
Intubation (%)	30.84 (29.90)	41
SpO2	85.61 (18.52)	13
Corticosteroids (%)	29.92 (21.99)	27
Comorbidity (%)	47.37 (27.63)	29
Cardio vascular disease	14.35 (10.03)	91
Hypertension (%)	36.61 (15.23)	103
Diabetes (%)	19.46 (11.29)	90
Obesity (%)	26.61 (16.69)	16
Fever (%)	81.23 (12.64)	73
Cough (%)	66.29 (17.33)	72
Dyspnea/Difficulty breathing (%)	42.05 (28.06)	58
Pharyngitis/sore throat/phyrangalgia(%)	13.79 (10.66)	35
Diarrhea (%)	15.04 (11.24)	55
Headache (%)	10.90 (9.82)	47
Lymphopenia (%)	58.08 (22.63)	11
WBC 103/mm3	13.50 (11.28)	78
ESR	35.62 (17.42)	15
LDH (U/L)	377.17 (150.91)	52
Elevated CRP (%)	50.54 (31.26)	11
Elevated LDH (%)	49.49 (22.27)	7
Elevated d‐dimer (%)	51.63 (21.82)	11
IL‐4 (pg/ml)	2.80 (1.62)	13
IL‐8 (pg/ml)	20.32 (13.05)	8
IL‐10 (pg/ml)	6.63 (5.06)	32

Abbreviations: ESR, erythrocyte sedimentation rate; ICU, intensive care unit; IL‐10, interleukin‐10; IL‐4, interleukin‐4; IL‐8, interleukin‐8; LDH, lactic dehydrogenase; SpO2, saturation of peripheral oxygen; WBC, white blood cells.

Data analysis procedure

This study examined the correlation between CRP and levels of various cytokines in patients with COVID‐19. First, we determined the aggregate mean and standard deviation (SD) of all the variables for which data were sought. The correlation between different Interleukins and CRP levels was measured using the correlation‐ coefficient (r) and p‐value. A significant relationship was defined as: p ≤ 0.05, 0.7 ≤ r ≤ 1. All the statistical analyses were conducted using SPSS software version 26.

RESULTS

Our initial search yielded 1307 records, of which 506 were duplicates and 801 records entered the title and abstract screening. Screening the title/abstract; 615 records were removed and 186 were eligible for full‐text screening. Finally, we found 103 eligible studies accounting for 101,309 total participants in the analysis (Figure 1).

The patients were mostly males older than 50 years suffering from a severe disease. Half of the participants had comorbidities, and the most prevalent one was hypertension. Half of the patients needed oxygen and one‐third needed intubation. Laboratory findings of the articles such as white blood cells (WBC), erythrocyte sedimentation rate (ESR), and IL‐10 were very heterogeneous and their standard deviations (SDs) were wide (Table 1).

This review examined the correlation between CRP levels and different cytokine levels in patients with COVID‐19. The combined mean and standard deviation (mean, SD) for laboratory and clinical presentations were as follows: LDH (377.17 U/L, SD: 150.91), WBC × 103/mm3 (13.50, SD: 11.29), SpO2 (85.61, SD: 18.52), fever (%) 81.23 (SD: 12.64), lymphopenia (58.08%, SD: 22.63), cough (66.29%, SD: 17.33) and oxygen supplementation(50.75%, SD: 34.74). Detailed information is illustrated in Table 2.

In Table 3, distribution of some variables based on median and interquartile range has also been reported.

Table 3

Distribution of some variables based on median and interquartile range in included studies

Characteristics	Median (IQR)	Number of studies
Asymptomatic (%)	0 (0)	30
Moderate (%)	10.0 (54.0)	53
Mortality (%)	10.0 (16.5)	75
COPD (%)	5.0 (9.5)	56
Autoimmune or inflammatory diseases (%)	1.0 (4.0)	20
Neutrophil count	5.0 (3.0)	57
Lymphocyte count	1.0 (0)	67
CRP (mg/dl)	10.39 (27.87)	103
Procalcitonin (ng/dl)	0 (5.0)	43
d‐dimer (microg/ml)	1.0 (2.0)	60
Elevated procalcitonin (%)	18.0 (46.0)	7
IL‐1 (pg/ml)	0.14 (0.42)	4
IL‐2 (pg/ml)	4.03 (5.23)	13
IL‐6 (pg/ml)	24.84 (52.06)	97
IL‐9 (pg/ml)	7.56 (22.13)	6
IL‐17 (pg/ml)	2.0 (38.21)	5

Abbreviations: COPD, chronic obstructive pulmonary disease; CRP, C‐reactive‐protein; IL, interleukin; IQR, interquartile range.

Table 4 shows the correlation between levels of different Interleukins and CRP levels using correlation‐coefficient (r) and p‐value. IL1, IL4, IL6, IL8, IL9, IL10, and IL17 were included in this review (Table 4). All of the interleukins examined in this study were associated with CRP levels; however, the most were insignificant. The analysis showed that IL10 have a moderate and significant relationship with CRP (p ≤ 0.05, r = 0.472). IL‐10 predicted almost 10% of CRP changes.

Table 4

Correlation‐coefficient (r) between Interleukins and CRP with Spearman index of correlation

Variable	IL‐1	IL‐4	IL‐6	IL‐8	IL‐9	IL‐10	IL‐17
C‐reactive protein	0.316*	0.413*	0.316**	0.321*	0.200*	0.472***	0.300*
Number of participants (n)	171	1872	15,479	4289	3919	6241	688
Number of studies (n)	4	13	97	8	6	32	5

p > 0.05,

p = 0.001

p = 0.020.

DISCUSSION

The available studies that were analyzed in this systematic review and analysis examined the relationship between CRP levels and different cytokines levels in people infected with the SARS‐CoV‐2 virus. The correlation analysis suggests that the CRP level has a moderate correlation with IL‐10 levels in COVID‐19 patients. However, there seems to be no significant correlation observed between CRP and IL‐1, IL‐6, IL‐8, IL‐9, and IL‐17 in aggregated data.

Inflammatory cytokines help to initiate the inflammatory response as well as regulate host defense against pathogens via the innate immune response. Pro‐inflammatory cytokines are signaling molecules secreted by immune cells such as T‐helper cells and macrophages, as well as certain other cell lines that promote inflammation. CRP on the other hand is related to the innate immune system that mostly indicates an inflammatory response. Principal findings of our analysis suggest that the level of various cytokines and CRP is markedly increased in COVID‐19 patients, especially in severe cases. However, the correlation between these interleukins and CRP was established for IL‐10 which showed a moderate relationship with CRP levels.

In COVID‐19 patients, inflammatory cytokines and tissue destruction stimulate CRP production. Critically ill COVID‐19 patients present with elevated CRP levels which are linked to an overproduction of inflammatory cytokines. The studies analyzed showed that COVID‐19 patients that were classified as severe cases usually have high CRP levels. Hence, it could be postulated that an elevated CRP level may be a useful early marker in predicting the possibility of disease progression. Various studies have indicated that CRP is a suitable indicator to determine the chance of disease progression from non‐severe to moderate and severe forms of the disease and even death. , , , , , , , Previous research found elevated levels of pro‐inflammatory cytokines and chemokines in SARS and MERS patients, indicating that excessive cytokine and chemokine responses performed a pathogenic role in CoV infection. This further enshrines the critical role of CRP in the prognosis of COVID‐19.

Aside from CRP being linked to COVID‐19 prognosis, numerous studies have investigated other immunological factors that could be implicated in this disease. Cytokine storm has been widely reported in this disease. It occurs by the high expression of IL‐6 and TNF‐α. , Patients with severe COVID‐19 have higher levels of IL‐2, IL‐6, IL‐7, IL‐8, IL‐10, IP‐10, MCP1, TNF, macrophage inflammatory protein 1 alpha (MIP1A), and granulocyte‐colony stimulating factor according to some studies (G‐CSF). , , , , , , , Recent data show that severe COVID‐19 causes a cytokine storm and are associated with poor clinical outcomes, with IL‐6 playing a key role. , Additionally, certain laboratory inspections, particularly lymphocytes, PCT, ALT, AST, LDH, d‐dimer, CD4 T cells, ferritin level, and IL‐6, could predict the progression of COVID‐19 changes, providing valuable signals for preventing disease deterioration. , , As such, there is a link discovered between IL‐6 levels and lymphocyte count, LDH, CRP, and procalcitonin. With high sensitivity and specificity, the optimal IL‐6 cutoff value has been suggested to be 30.95 pg/ml. , , , , , , , , , Interestingly, elevated serum levels of CRP and lactate dehydrogenase (LDH) seem to exhibit a significant reduction after antiviral therapy but serum levels of proinflammatory markers like IL‐6, IL‐8, and IFNγ remain elevated at baseline in COVID‐19 patients.

Furthermore, dynamic cytokine storms as well as high serum IL‐6 and TNF‐ serum levels, were found to be independent and significant predictors of disease severity and death. , , , , , , , , , , , , , , , , , , Conversely, some interleukins have been reported by studies indicating that they are not linked to the prognosis of this disease. IL‐4, IL‐10, IL‐1β, and IL‐17 were found by various studies to have less impact on the outcome of COVID‐19. , However, the analysis from this study suggests a moderate correlation between IL‐10 and CRP levels in COVID‐19 patients. Using the Spearman index of correlation in this study, postulates IL‐10 to have a mild to moderate correlation with CRP levels than the other interleukins included in the current study. Our findings are further strengthened since previous explorations by Del Valle‐Mendoza et al. and Zhang et al. established a relationship between IL‐10 and COVID‐19 severity. However, the latter in their systematic review and meta‐analysis only showed the relationship between disease severity and cytokines but not CRP levels.

Models designed to predict the course of the disease concerning inflammatory markers suggest a good predictive role of IL‐6 and IL‐2 in determining disease progression. However, the addition of more proinflammatory markers weakens the predictive strength of these models. Notwithstanding, CRP, white blood cell count, IL‐6, and IL‐10 have been suggested to be significantly associated with severe disease and mortality in a meta‐analysis of 8719 SARS‐CoV‐2 infected patients. Soltani‐Zangbar et al. in their study also established the crucial role of vitamin D in the fight against the cytokine storm in COVID‐19 patients, further strengthening the multifaceted nature of inflammatory cytokines’ role in this disease. As a result, vitamin D supplementation may be considered an adjunct therapy for COVID‐19 patients. Additionally, glucosinolates and glucosinolate hydrolysis products as valuable nutraceuticals, have been suggested to be helpful in the prevention and treatment of COVID‐19‐related cytokine storms through their anti‐inflammatory and antioxidant activities.

Researchers have attempted to estimate the cut‐off reference values for estimating the COVID‐19 disease course since biomarker testing like interleukins and CRP could help in establishing a scoring method for the prediction and guide for appropriate treatment. Though more large‐scale studies on COVID‐19 patients are needed to build precise threshold values, a cutoff of 30.95 pg/ml has been suggested specifically for IL‐6. , This could predict the progression of the infection from mild to severe. The measurement of immune cells and cytokines may help identify immune markers of COVID‐19 severity and contribute to the development of immunologic therapies and vaccines, as the latter is considered a potent way of curbing the pandemic over the past years and even future pandemics. , , While laboratory testing and investigations are evolving at a tremendous pace, rapid policies and clinical guidelines are essential for the early screening and stratification of patients as well as lasting solutions in terms of health and governmental policies to manage future outbreaks aptly. ,

The current study identified higher levels of WBC, ESR, CRP, and IL‐10 as inflammatory markers that are correlated with the severity of COVID‐19, hence, measuring these markers may speed early recognition of the disease progression and could be significant in their clinical treatments.

Strengths, limitations, and recommendations

This systematic review and correlation analysis have some strength. All important databases were searched and included all published articles to minimize the risk of publication bias. As a result, this study presents a wide range of study participants, over one hundred thousand, representing significant global coverage. The Spearman index of correlation was used to determine the relationship between CRP and a wide range of interleukins including IL‐1, IL‐4, IL‐6, IL‐8, IL‐9, IL‐10, and IL‐17. Hence, the current study presents an added advantage over previous studies that tackled a limited number of interleukins and barely their relationship with CRP levels in COVID‐19 patients.

Several limitations of our study should be considered. The majority of the included studies in this correlation analysis were retrospectives. Also, the overall generalize ability of the systematic review and correlation results should be interpreted with caution as most of the included studies were conducted in China due to limitations in geographic distribution and ethnic diversity, though there were a few from other countries. It would be better to include more studies with a broad geographic scope to gain a more comprehensive understanding of the CRP and interleukin correlation in COVID‐19 patients. Consequently, more prospective studies are recommended in this regard to aid with the better application of the findings in the clinical treatment of these patients and probably future pandemics of similar pathophysiology.

CONCLUSION

This correlation analysis suggests IL‐10 is correlated with CRP in patients with COVID‐19 infection, indicating that, patients with elevated levels of IL‐10 may progress to a severe form of the disease. Although other interleukins (IL‐1, IL‐4, IL‐6, IL‐8, IL‐9, and IL‐17) were also associated with CRP, their association was not significant compared to IL‐10. A better understanding of the pro‐inflammatory markers could contribute to the implementation of therapeutic and preventive approaches. More prospective studies are suggested to explore the relationship between CRP and cytokines as potential markers for the early identification of COVID‐19 progression and severity.

AUTHOR CONTRIBUTIONS

The conception and design of the study: Esmaeil Mehraeen and SeyedAhmad SeyedAlinaghi. Acquisition of data: Amirali Karimi, Pegah Mirzapour. Analysis and interpretation of data: SeyedAhmad SeyedAlinaghi, Fatemeh Afroughi, Alireza Noroozi, and Ghazal Arjmand. Drafting the article: Shayan Abshenas, Zahra Pashaei, Marcarious M. Tantuoyir, Omid Dadras, Kowsar Qaderi, Solmaz Saeidi, Soheil Dehghani, Ayda Shabanzadeh Pirsaraie, and Amir Masoud Afsahi. Revising it critically for important intellectual content: SeyedAhmad SeyedAlinaghi Esmaeil Mehraeen. Final approval of the version to be submitted: Esmaeil Mehraeen, Omid Dadras, and SeyedAhmad SeyedAlinaghi.

CONFLICTS OF INTEREST

The authors declare no conflicts of interest.

TRANSPARENCY STATEMENT

The lead author Esmaeil Mehraeen affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Supporting information.

Click here for additional data file.