key: cord-0806596-rgtdk52u
authors: Huang, Rong; Xie, Liangcai; He, Junpeng; Dong, Hong; Liu, Tianchun
title: Association between the peripheral blood eosinophil counts and COVID-19: A meta-analysis
date: 2021-06-11
journal: Medicine (Baltimore)
DOI: 10.1097/md.0000000000026047
sha: 891baebbf0dc2e12618a3be2790912485cdd460d
doc_id: 806596
cord_uid: rgtdk52u

BACKGROUND: The conclusions about the relationship between eosinophil counts and the severity of coronavirus disease 2019 (COVID-19) were controversial, so we updated the evidences and reassessed it. METHODS: We searched the PubMed, Cochrane library, Excerpta Medica Database, and Web of Science to compare the eosinophil counts about non-severe disease group (mild pneumonia, moderate pneumonia, non-critical disease and recovery group) and severe disease group (severe pneumonia, critical pneumonia, critical disease and death group) in COVID-19. RESULTS: A total of 1228 patients from 10 studies were included. Compared with non-severe group, severe group had strikingly lower average eosinophil counts (SMD 0.65, 95% confidence intervals [CI] 0.29–1.01; P < .001). The result of subgroup analysis of different countries showed SMD 0.66, 95% CI 0.26–1.06; P < .001. Another subgroup analysis between mild-moderate pneumonia versus severe-critical pneumonia showed SMD 0.69, 95% CI 0.25–1.13; P < .001, and no significant risk of publication bias (Begg test 0.063 and Egger test 0.057) in this subgroup. The heterogeneity was substantial, but the sensitivity analyses showed no significant change when individual study was excluded, which suggested the crediblity and stablity of our results. CONCLUSIONS: The eosinophil counts had important value as an indicator of severity in patients with COVID-19. PROSPERO REGISTRATION NUMBER: CRD42020205497.

A new form of respiratory and systemic disorder named coronavirus disease 2019 is caused by the severe acute respiratory syndrome coronavirus 2, [1, 2] a new and more infectious virus than that of severe acute respiratory syndrome and Middle East respiratory syndrome. The clinical symptoms of COVID-19 are fever, cough, fatigue, extremity pain, and gastrointestinal symptoms. [3] Severe patients often experience progressive dyspnea and/or refractory hypoxemia 1 week after onset, and rapidly develop acute respiratory distress syndrome, septic shock, multiple organ failure and other manifestations. [4, 5] There is no doubt that it has become one of the most concerned issues today because of its extremely high transmission capacity and mortality rate. According to Guidelines for the Diagnosis and Treatment of COVID-19, the clinical typing of COVID-19 are 4 types: mild, moderate, severe and critical pneumonia. So far, over 64 million people has infected and caused over 1,480,000 deaths (as of December 2, 2020) over the world. Since it breaks out, the global economy is badly disrupted and global health care systems is suffered a multitude of challenges.

One of the first lines to against virus, the eosinophil should be noticed. [6] A previous systematic literature included 3 studies showed that the peripheral blood eosinophil counts may not be associated with the progression of COVID-19. [7] However, several related clinical studies had been conducted since then, these findings were not entirely consistent. Yang et al thought the low percentage of eosinophil was not a biomarker of pneumonia severity. [8] While more scholars got the conclusions that the eosinophil counts were helpful to predict the severe COVID-19 cases. [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] Considering these divergent conclusions, it is worth exploring whether eosinopenia is related to the severity of the disease in COVID-19, so we meta-analyze the relevant literature. 

We performed electronic searches in PubMed, Cochrane Library, EMBASE and Web of Science to search published studies without date (until 23 Aug, 2020) or language restrictions, using the following terms: (a) "Eosinophils" OR "White Blood Cells" OR "Leukocytes" OR "Granulocytes"; combined with (b) "2019 nCoV" OR "2019 novel coronavirus" OR "new coronavirus" Studies of other groups (11) No useable counƟng data (2) No value data to calculate SMD (2) Studies included in qualitaƟve synthesis (n = 10)

Studies included in quanƟtaƟve synthesis (meta-analysis) (n = 10) OR "novel coronavirus" OR "novel corona virus" OR "SARS CoV-2" OR "Wuhan corona virus " OR "COVID-19" without date (until August 23, 2020) or language restrictions. Searches were re-run prior to the final analysis. The flow chart of search result was presented in Figure 1 .

The articles found in the search were imported into Endnote X9 software. Duplicates were eliminated. Two investigators (RH and LCX) independently screened the titles and abstracts due to the inclusion and exclusion criteria, the irrelevant articles were removed. Finally, the eligible articles were chosen by full-text screening. Disagreements were resolved by discussion between these 2 authors and the third author (JPH) resolved any disagreements.

For all the eligible studies, 2 reviewers (RH and LCX) independently extracted the information: the first author, year, study country, demographic information (age and sex), disease severity, study size and outcomes. If the data given in the original article was nonconsistent forms, we would use Wan's calculation formula to convert the data to mean ± SD. [21] 

Two reviewers (RH and LCX) independently took the Newcastle-Ottawa Quality Assessment Scale [22] for quality estimation of literature. Nine questions would be evaluated, and each satisfactory answer were received 1 point, the maximum score was 9. If the score ≥7, we considered the article was high methodological quality. Two reviewers (RH and LCX) crosschecked the assessing process and the final score were obtained after discussing with the third reviewer (JPH).

1. Selection: Representativeness of the exposed cohort; Selection of the non-exposed cohort; Ascertainment of exposure (the proof of COVID-19 and the eosinophil counts measurement); Demonstration that outcome of interest was not present at start of study. 2. Comparability: Comparability of cohorts on the basis of the design or analysis. (a) Study controls for the eosinophil counts;

(b) Study controls for any additional factor (age, gender, exposure history, comorbidity, etc). 3. Outcome: Assessment of outcome; was follow-up long enough for outcomes to occur? (death or recurrence); Adequacy of follow-up of cohorts.

2.6. Statistical analysis 2.6.1. Assessment of heterogeneity and data syntheses. The standard mean differences and corresponding 95% confidence intervals (CIs) were calculated. Heterogeneity presumption was investigated using the Q-test and I 2 statistic. If there was significant heterogeneity among studies (P < .05 or I 2 > 50%), a random-effects model was used; otherwise, a fixed-effects mode was applied. Two-sided P < .05 was considered statistically significant. Analyses were conducted using Stata 15.0 statistical software packages.

2.6.2. Subgroup and sensitivity analyses. If there was significant heterogeneity among studies, the subgroup analysis of different countries and another subgroup analysis between mild-moderate pneumonia vs severe-critical pneumonia were assessed. Sensitivity analysis were performed by sequentially omitting each trial. 

At the beginning, 12 articles were finally eligible according to the search strategies, [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] but 2 articles were unable to extract value data to calculate SMD [14, 18] and finally we included 10 articles assessed 1228 patients considered in this meta-analysis. The sample size ranged from 9 to 231. All the studies were originated from Asia. The details of these articles included in the metaanalysis were summarized in Table 1 .

A comparison of the peripheral blood eosinophil counts for nonsevere and severe disease in COVID-19 was performed in 10 Table 1 Basic characteristics of included studies. studies. Compared with non-severe disease group, severe disease group had strikingly lower average eosinophil counts (SMD 0.65, 95% CI 0.29-1.01; P < .001, I 2 = 83.7%) (Fig. 2) . Considering the high heterogeneity, 2 subgroups were studied. Between different countries, the result showed SMD 0.66, 95% CI 0.26-1.06; P < .001, I 2 = 85.2% (Fig. 3) . Another subgroup was according to the strict definition of clinical classification about COVID-19 (mild-moderate pneumonia vs severe-critical pneumonia), the result showed SMD 0.69, 95% CI 0.25-1.13; P < .001, I 2 = 87.1% (Fig. 4) . The sensitivity analyses showed no significant change when individual study was excluded using random-effects model, which suggested the crediblity and stablity of our results (Fig. 5) . Begg funnel plot and Egger test were used to assess publication bias in the interested subgroup, the Begg funnel plot showed no evidence of significant publication bias (P = .063) nor did Egger (P = .057) (Fig. 6 ).

In the present study, we updated the evidences and observed that the peripheral blood eosinophil counts levels in patients with severe disease group were significantly lower than those with non-severe disease group in COVID-19. On the basis of this meta-analysis, we suggest that the eosinophil counts levels have important value as an indicator of severity in patients with COVID-19. It may also be used to monitor the progression of this disease as early as possible. Despite eosinophils are the second least granulocyte subpopulation in the peripheral blood, the scientific community is getting increasingly interested in it because of its complex pathophysiological roles to against bacterial and viral pathogens. Evidences are showing that eosinophils protect body from viral infections, especially against RNA viruses. [23] The TLR7 receptor, which recognizes viruses and recognizes single stranded RNA (ssRNA), is one of the most important viral receptors in eosinophils, demonstrating the role of eosinophils in virus recognition is unarguable. [6] COVID-19 is becoming one of the worst infection disease outbreaks known, it is transmitted by droplets, aerosol, contact and fecal-oral route infection, and tend to transmit among family clusters [24, 25] or cause outbreaks in hospitals. It is important to identify the key clinical features of COVID-19 patients, which Medicine may help to detect and isolate the infected individuals as early as possible, and minimize the spread of the disease. Some scholars had studied the cases of COVID-19 and showed that the severe disease had abnormalities in many laboratory parameters, and some of them can be used as predictors of disease severity, such as levels of lactate dehydrogenase, [26, 27] lymphocyte and subset counts, [27] [28] [29] [30] [31] interleukin-6, [32, 33] procalcitonin, [34] D-dimer, [35] C-reactive protein [27, 33] and so on. But compared to other biomarkers, the eosinophil was often overlooked, it had long been thought to be associated with allergy diseases and parasitic infections traditionally. Actually, it was versatile cell, and some scholars believed it was positioned centrally within immune and inflammatory networks, its new roles as neoplasm surveillance, tissue remodeling and the restructuring of adipose tissue were emerging. [36] The eosinophil cells may act as a positive predictor in early stages during the coronavirus infection. [37] It has been reported previously that in infected patients, increased inflammation leads to lower counts of eosinophils. [38] Increased count of eosinophils was associated with a better prognosis for COVID-19, including the lower incidence of complications and mortality. While the recovery of lymphocytes had no effect on the prognosis. [39] In surviving acute lung injury patients, the number of eosinophils in the lungs showed an increased compared to the non-survivors. [40] Similarly, the number of peripheral blood eosinophils also increased in surviving acute respiratory distress syndrome patients. Therefore, the scholars further investigated its molecular mechanism and found that in the initiation of acute lung injury, CD101 À eosinophils increased more rapidly and briefly than the neutrophils and secreted Protectin-D1 through Alox15-mediated to reduce the accumulation of inflammatory cells and reduce inflammatory factors, thus playing a role in fighting lung inflammation. [41] Whether the glucocorticoids can save the lives of patients with severe in COVID-19 is still controversial. The hormones can prevent the release of eosinophils in the bone marrow and reduce them in the peripheral blood. Given that the eosinophils play a "protective role" in the onset of COVID-19, early use of hormones will quickly inhibit their proliferation, differentiation, migration and "protective effect" against diseases.

Several limitations of our study should be discussed. Firstly, the number of studies included was relatively low although the definition of clinical classification about COVID-19 was not strict, which may reduce the reliability of results. Secondly, the data collection may be insufficient because only English language papers were included. Thirdly, because of the small sample size, it may not strongly confirm the association between eosinophil counts levels and severity of COVID-19, which needed more studies. Fourthly, the studies included in this analysis were all performed in Asia, and the results may not be representative of other parts of the world. The significant heterogeneity in these studies could not be neglected. The possible reasons for the observed heterogeneity were the difference in underlying comorbidities, different patient population, co-infection with other diseases and the variation in follow-up. The looser inclusion criteria resulted the variety of control interventions, which may also lead to heterogeneity. However, when we chose the random-effects model and performed subgroup analysis according to the strict definition, the marked heterogeneity was still remained.

In spite of the limitations mentioned above, the subgroup analysis and its sensitivity were remained consistent, demonstrating that eosinophil counts levels might be useful as an important parameter to recognize patients with severe COVID-19 in the disease course. If eosinophils are activated, they may have the potential to treat viral respiratory diseases. [42] We believe this analysis will contribute to the development of anti-COVID-19 drugs. 

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Novel Wuhan (2019-nCoV) Coronavirus. American journal of respiratory and critical care medicine 2020

Quinolines-based SARS-CoV-2 3CLpro and RdRp inhibitors and spike-RBD-ACE2 inhibitor for drug-repurposing against COVID-19: an in silico analysis

Risk factors associated with acute respiratory distress syndrome anddeath in patients with coronavirus disease 2019 pneumonia in Wuhan

Risk factors associated with disease progression in a cohort of patients infected with the 2019 novel coronavirus

Eosinophil response against classical and emerging respiratory viruses: COVID-19

Eosinophil count in severe coronavirus disease 2019

Clinical characteristics and eosinophils in young SARS-CoV-2-Positive Chinese travelers returning to Shanghai

Abnormalities of peripheral blood system in patients with COVID-19 in Wenzhou, China

Characteristics and prognostic factors of disease severity in patients with COVID-19: the Beijing experience

Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan

Clinical features of 69 cases with coronavirus disease 2019 in Wuhan

Distinguishing between COVID-19 and influenza during the early stages by measurement of peripheral blood parameters

Dysregulation of immune response in patients with coronavirus 2019 (COVID-19) in Wuhan

Epidemiologic and clinical characteristics of 91 hospitalized patients with COVID-19 in Zhejiang, China: a retrospective, multi-centre case series

Haematological characteristics and risk factors in the classification and prognosis evaluation of COVID-19: a retrospective cohort study

Preliminary study to identify severe from moderate cases of COVID-19 using combined hematology parameters

The role of peripheral blood eosinophil counts in COVID-19 patients

The underlying changes and predicting role of peripheral blood inflammatory cells in severe COVID-19 patients: a sentinel?

Clinical profiles, characteristics, and outcomes of the first 100 admitted COVID-19 patients in Pakistan: a single-center retrospective study in a tertiary care hospital of Karachi

Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range

Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses

The role of innate leukocytes during influenza virus infection

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster

Lactate dehydrogenase elevations is associated with severity of COVID-19: a meta-analysis

Laboratory findings of COVID-19: a systematic review and meta-analysis

Hematologic, biochemical and immune biomarker abnormalities associated with severe illness and mortality in coronavirus disease 2019 (COVID-19): a meta-analysis

CD4+T, CD8+T counts and severe COVID-19: a metaanalysis

The laboratory tests and host immunity of COVID-19 patients with different severity of illness

Characteristics of peripheral lymphocyte subset alteration in COVID-19 pneumonia

Interleukin-6 as a potential biomarker of COVID-19 progression

Prognostic value of interleukin-6, C-reactive protein, and procalcitonin in patients with COVID-19

Procalcitonin in patients with severe coronavirus disease 2019 (COVID-19): a meta-analysis

D-dimer levels on admission to predict in-hospital mortality in patients with Covid-19

Eosinophils: friends or foes

Complete blood count might help to identify subjects with high probability of testing positive to SARS-CoV-2

A comparison of eosinopenia and C-reactive protein as a marker of bloodstream infections in critically ill patients: a case control study

Immunodetection of occult eosinophils in lung tissue biopsies may help predict survival in acute lung injury

The prognostic value of eosinophil recovery in COVID-19: a multicentre, retrospective cohort study on patients hospitalised in Spanish hospitals

Homeostatic and early-recruited CD101 -eosinophils suppress endotoxin-induced acute lung injury

Efficacy, adverse events, and inter-drug comparison of mepolizumab and reslizumab anti-IL-5 treatments of severe asthmaa systematic review and meta-analysis

The authors have no funding and conflicts of interests to disclose.Data sharing not applicable to this article as no datasets were generated or analyzed during the current study. 

Writingoriginal draft: Rong Huang, Liangcai Xie, Junpeng He, Hong Dong, Tianchun Liu.