Mean platelet volume/platelet count ratio predicts severe pneumonia of COVID‐19

Abstract Background Although platelet mean volume/platelet count ratio (MPR) is considered to be a crucial marker of inflammatory and infectious diseases, the relationship between MPR and novel coronavirus infectious disease 2019 (COVID‐19) remains unclear. Methods In this retrospective study, 85 patients with confirmed COVID‐19 were enrolled and divided into low and high MPR group. Data from repeated measures were compared by the generalized estimating equations. Cox regression analyses were performed to assess the impact of MPR on the incidence of severe pneumonia (SP), with inverse probability of treatment weighting (IPTW) used to reduce confounding bias. The primary outcome is the incidence of SP of COVID‐19. Results During follow‐up, 17 (20.0%) patients were developed to SP. Compared with mild patients, patients with SP developed showed a higher MPR level at baseline, day 1, day 2, and day 3 after admission (P = .005, P = .015, P = .009, and P = .032, respectively). Kaplan‐Meier method showed a higher incidence of SP in the high MPR group than the low MPR group (log‐rank test = 10.66, P = .001). After adjustment, high MPR was associated with an elevated incidence of SP (HR, 5.841, 95% CI, 1.566‐21.791, P = .009). The IPTW method also suggested that MPR was a significant factor related to the incidence of SP (HR, 8.337, 95% CI, 4.045‐17.182, P < .001). Conclusion High MPR level is an independent risk factor for severe pneumonia in patients with COVID‐19.

expectoration, fatigue, upper respiratory congestion, ecphysesis, gastrointestinal symptoms, and myalgia/arthralgia. 4,5 Most patients present mild symptoms, but it may progress to severe symptoms in some patients (especially the elderly and/or patients with comorbidity). 4 Severe patients develop rapidly into acute respiratory failure, acute respiratory distress syndrome (ARDS), septic shock, metabolic acidosis, and coagulation disorders. 6 The prognosis of patients with severe pneumonia (SP) of COVID-19 is worse than that of patients with mild type. Early recognization of the risk factors of SP contributes to antecedent intervention, which shows crucial clinical significance for the treatment and prognosis of patients. SP associates with plenty of abnormal laborator y tests, which indicate different injuries of specific organs. 7 Complex pathophysiology involves infection and systemic inflammator y response, causing a series of reactions such as coagulation activation, liver damage, myocardial damage, and kidney damage. 8 Platelets play a significant role in the procedure of inflammation and coagulation, activated platelets release a great number of substances, which belong to the key factors of inflammation. 9 Mean platelet volume (MPV) has been regarded as a surrogate marker of platelet activation. 10 MPV was a useful prognostic indicator for critical patients. [11][12][13] Besides, some studies have shown a negative correlation between MPV and platelet count (PC) in severe patients. 14,15 It is reported that the combination of MPV and PC could be more clinically significant than MPV or PC alone. 14,16 As far as we know, no published article investigated the relationship between MPV/platelet count ratio (MPR) and the prognosis of COVID-19. Our research aims to assess whether MPR is a useful predictor for the progression of COVID-19.

| Study participants
From January 21, 2020, to February 14, 2020, patients who suffered from COVID-19 were recruited from two hospitals in southern China. Patients who were younger than 18 or suffered from liver diseases were excluded from this study. This retrospective cohort study was approved by the Ethics Committee. Written informed consent was waived.

| Definitions
COVID-19 was diagnosed and classified based on the guidance from the World Health Organization (WHO). 17 SP of COVID-19 was defined as fever or suspected respiratory infection, plus one of the following: (a) respiratory rate >30 breaths/min, (b) severe respiratory distress, and (c) SpO2 ≤ 93% on room air.

| Data collection
Epidemiological, clinical, laboratory, radiological data were collected from case report forms and electronic medical records. The information recorded included demographic data, comorbid conditions, exposure history, signs, laboratory data, chest computed tomographic (CT) scans, and treatments. Laboratory findings, Chest CT results, and treatment measures during the hospital stay were monitored.
Baseline laboratory data were collected on admission to the hospital.
Samples of peripheral blood were collected into tubes with ethylenediamine tetraacetic acid. Mean platelet volume and platelet distribution width were determined using an automated blood cell analyzer (XN-2000, Sysmex) within 30 minutes of sample collection. MPR was defined as mean platelet volume (fL)/platelet count (^10 9 /L) * 100%.
Health-care providers or responsible doctors were contacted for clarification if data were missing or not clear from records.

| Follow-up
After admission, patients were re-examined for laboratory data and CT if necessary. Clinical outcomes were monitored up to February 19, 2020, the final date of follow-up. Patients were censored if they were transferred, discharged, or still hospitalized without SP until the final date of follow-up. The primary outcome of this study was the SP developed.

| Statistical analysis
Patients were divided into low and high MPR group according to the optimal value of baseline MPR for predicting SP developed.
Baseline data were expressed as mean ± standard deviation or median (interquartile range) for continuous data, and number (%) for categorical data. Differences between two groups were compared using Student's t test or Mann-Whitney U test for continuous data, and chi-square test, Fisher's exact test, or Kruskal-Wallis test for categorical data. The receiver operating characteristic (ROC) curve was performed to assess the predictive ability of baseline MPR for the incidence of SP and to acquire MPR cutoff values to maximize sensitivity and specificity. Spearman's correlation coefficients were used to assess the relationships between MPR and other clinical variables. The generalized estimating equations were conducted to compare data from repeated measures. Kaplan-Meier method was used to estimate the incidence of SP, and differences between two groups were compared by the log-rank test. The relationships of SP developed and MPR were determined by univariate and backward stepwise multivariate Cox regression analysis.
Age, sex, center, and variables with P < .1 in univariate analysis were entered into the multivariate model. In Cox models, time at risk was from study entry till SP developed, discharge, transfer, or the final date of follow-up. To reduce the influence of unbalanced distribution of confounding factors, the weighted Cox regression analysis was used to adjust confounders through the inverse probability of treatment weighting (IPTW). The propensity score was calculated in each case by a logistic regression model including factors in the Cox model. Remaining missing data were filled using the missForest R package. All statistical analyses were performed using SPSS (Statistical Package for the Social Sciences) version 25.0 software (SPSS Inc) and R software (version R-3.5.5, www.rproje ct.org). P < .05 were considered statistically significant.

| Baseline characteristics
By February 14, 2020, 101 patients with confirmed COVID-19 had been admitted, of whom 16 (12.9%) were considered ineligible due to age < 18 or liver diseases. 85 patients were ultimately enrolled in this study ( Figure 1). The demographics and characteristics data of two groups, categorized based on the optimal cutoff value of baseline MPR for predicting SP developed, were presented in Table 1.
The median days from symptom onset to admission were 3.0 (2.0, 5.0) days. No patients had been exposed to the Wuhan seafood market but 28 (32.9%) had made short-term trips to Wuhan city before illness onset. Patients in the high MPR group showed an increased prevalence of hypertension (26.9% vs 6.8%, P = .030), but no significant difference was observed in other comorbid conditions between two groups. Among patients with SP developed, the median days from admission to SP were 6.0 (5.0, 11.5) days. Among There were significant differences in laboratory findings on admission between two MPR levels. Patients in high MPR group showed lower white blood cell, lymphocyte, platelet, and albumin (P = .021, P = 0 0.031, P < .001, and P = .010, respectively), and higher MPV, PDW, creatinine, and hs-CRP (P < .001, P = .004, P = .030, and P = .006, respectively).

| Correlations between baseline MPR and clinical variables
The correlations between baseline MPR and clinical variables were presented in Table 2 using the Spearman correlation coefficient.

| Dynamic profile of MPR in patients with COVID-19
To compare the dynamic changes of MPR between mild and severe patients in the early stage, the MPR data were tracked from day 0 to day 7 after admission at one-day intervals (Figure 3)

| MPR associated with SP developed
Kaplan-Meier curves for MPR level based on the optimal cutoff value were shown in Figure 4. Patients in the high MPR group had an obviously higher incidence of SP than those in the low MPR group during the course of their hospitalization (log-rank test χ 2 = 10.66, P = .001).
The relationship of MPR with SP incidence with defined models (with low MPR group as the reference group) was shown in

| D ISCUSS I ON
In this study, we found that even after adjusting for multiple confounding factors, the increased baseline MPR was apparently associ- Platelets are essential for inflammation, hemostasis, and immune regulation. 19,20 Platelet activation plays a crucial role in viral pneumonia. 21 In viral pneumonia, platelet activation may cause lung injury by stimulating the detrimental inflammatory response of respiratory tract. 22 Activated platelets not only release inflammatory mediators but also expose surface molecules (including E-selectin and P-selectin), which affect the interaction between platelets and other cells, causing systemic inflammation and immune response. 9 The decrease in PC is associated with poor prognosis and increased mortality in hospitalized patients. 23,24 In contrast, PC elevation is bound up with a good prognosis. 25 Besides, for pneumonia patients, thrombocytopenia is a sign of adverse prognosis in hospitalized pneumonia patients and intensive care unit (ICU) patients hospitalized with severe community-acquired pneumonia (CAP). 26,27 Previous studies have revealed that the increase of MPV is relevant to the adverse prognoses of patients, such as CVD, chronic inflammatory disease, severe pneumonia, septic shock, and other diseases. [27][28][29][30][31] In patients with severe pneumonia, the increase of MPV after admission may be able to forecast the mortality. 32,33 MPV is negatively correlated with PC in severe patients. 15 MPR, which seems to be a better indicator of platelet function, has been proposed. 12,14,20,30,31,34 In many diseases, the increase of MPR is obviously associated with adverse prognoses, such as pneumonia after ischemic stroke, sepsis, critical disease, CVD, febrile epilepsy in children, and malignant tumor. 15,20,34,35 Ranias et al have shown that if pneumonia occurs in patients with ischemic stroke, an increase in the MPV/PC ratio could predict 30 days mortality. 35 Our study suggested that the increase of MPR was also a predictor of poor prognosis in patients with pneumonia caused by COVID-19.
The pathophysiological mechanism of MPR in predicting the adverse prognosis of patients with COVID-19 is not clear, but it may be related to the following mechanisms. Firstly, under the condition of inflammation, platelet production will increase owing to the increased synthesis of thrombopoietin mediated by multifarious cytokines. 22,36,37 Secondly, MPV reflects the metabolism and proliferation of megakaryocytes and platelet production in bone marrow. 9,13,21 In the beginning, when infection occurs, the release of many inflammatory cytokines (such as interleukin-1 (IL-1), IL-3 and IL-6, and tumor necrosis factor-α (TNF-α)) increases, leading to the increase of thrombopoietin and the expression of young platelets in the blood stream, 9,32 which causes the increase of MPV. Besides, after stress-induced platelet destruction, the decrease of PC further stimulates megakaryocyte to produce a large number of platelets, which also leads to an increase of MPV. 38  risk of poor outcomes in this situation. Further large-scale prospective studies are needed to confirm the association between MPR and COVID-19.
In conclusion, our study suggested that MPR is a useful indicator to help predict whether COVID-19 patients will progress to severe pneumonia. The early application of MPR is conducive to the hierarchical management of patients' risks and alleviates the shortage of medical resources.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no financial conflicts of interest.
Human and animal rights All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee at which the studies were conducted (IRB approval number 2020-hg-ks-04), and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

ACK N OWLED G EM ENTS
We are very grateful to Dr Yihao Li, Dr Liya Zhu, Dr Huaxing Huang and Dr Yueqiang Wen for helpful discussions and suggestions.