Hematological manifestations of COVID-19 acute respiratory distress syndrome patients and the impact of thrombocytopenia on disease outcomes: A retrospective study.

Background: Platelet count is a simple and readily available biomarker, in which thrombocytopenia was shown to be independently associated with disease severity and risk of mortality in the critical coronavirus disease-19 (COVID-19) patients. The aim of this study was to investigate the impact of thrombocytopenia on disease progression in critically ill COVID-19 patients with acute respiratory distress syndrome (ARDS) admitted to a medical intensive care unit (ICU).
Methods: COVID-19-associated ARDS patients in our research hospitals' ICU were retrospectively investigated. Patients were divided into two groups as thrombocytopenic (<150 × 109/ml) patients on admission or those who developed thrombocytopenia during ICU follow-up (Group 1) and those without thrombocytopenia during ICU course and follow-up (Group 2).
Results: The median platelet count of all patients was 240 × 109/ml, and the median D-dimer was 1.16 mg/ml. On admission, 32 (18.3%) patients had thrombocytopenia. The mean platelet count of Group 1 was 100.0 ± 47.5 × 109/ml. Group 1 was older and their Acute Physiology and Chronic Health Evaluation II and sequential organ failure assessment scores were higher. Group 1 had lower hemoglobin, neutrophil, and lymphocyte counts and higher ferritin and procalcitonin level. Invasive mechanical ventilation was more commonly needed, and disseminated intravascular coagulation (DIC) was more frequently observed in Group 1. The ICU and hospital length of stay of Group 1 was longer with higher mortality.
Conclusion: Patients with thrombocytopenia had increased inflammatory markers, frequency of DIC, duration of ICU stay, and mortality. The presence of thrombocytopenia may reflect the progression of COVID-19 toward an unfavorable outcome. Copyright:

INTRODUCTION

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) causing coronavirus disease-2019 (COVID-19) has rapidly emerged as a pandemic outbreak. As of November 4, 2021, there have been 251,266,207 confirmed cases of COVID-19, including 5,070,244 deaths reported by WHO.[1] Roughly 20% of patients require hospitalization, with an associated mortality of 11% among admitted patients in the United States.[2] Severe COVID-19 can lead to a critical illness with acute respiratory distress syndrome (ARDS) as its primary complication, and eventually, it can be followed by multi-organ failure and hypercoagulopathy.[3] Today, COVID-19-associated critical illnesses are not limited to respiratory manifestations that culminate in ARDS but also include other systems such as cardiovascular, renal, gastrointestinal, and hematopoietic.[3] Severe COVID-19 mostly presents with abnormal laboratory markers such as elevated C-reactive protein (CRP) and lactate dehydrogenase, and abnormal coagulation parameters, which may indicate unfavorable clinical outcomes. Critical COVID-19 patients are commonly complicated by lymphopenia, thrombocytopenia, coagulopathy, which often progresses to disseminated intravascular coagulation (DIC).[4] At the microvascular level, endothelial damage triggered by a systemic inflammatory response leads to hyperactivation of platelets and thrombosis causing disseminated platelet consumption.[567]

Platelet count is a simple and readily available biomarker, in which thrombocytopenia was shown to be independently associated with disease severity and risk of mortality in critically ill COVID-19 patients.[89] Thrombocytopenia has been used as a reliable biomarker in numerous scoring scales such as the sequential organ failure assessment (SOFA) score.[10] Adding to the complexity and diversity of COVID-19 clinical manifestations, this study aimed to investigate the hematological manifestations in COVID-19 patients with ARDS on intensive care unit (ICU) admission and the impact of thrombocytopenia on the disease prognosis. The primary outcome of the study was to investigate the impact of thrombocytopenia in the prognosis of COVID-19 patients, while the secondary outcome was to investigate whether or not the clinical and laboratory parameters of the thrombocytopenic group differ significantly from patients with a normal platelet count.

METHODS

Clinical and laboratory data of the ARDS patients who were treated in the COVID-19 ICU of our research hospital between July 1, 2020 and October 5, 2020, were retrospectively investigated. Inclusion criteria were as follows: patients over the age of 18 years, who met the Berlin 2012 ARDS criteria during COVID-19 ICU hospitalization, did not have a known hematological disease (myelodysplastic syndrome, myeloproliferative disease, etc.), had no active malignancy, and those had not received chemotherapy or radiotherapy in the last 6 months were included in the study. Patients with pseudo-thrombocytopenia in the peripheral smear were excluded from the study. In addition, patients whose severity of COVID-19 could not be associated with the clinical course of the disease, and thought to be immune related, or who were positive for any immune markers (antiphospholipid antibodies: lupus anticoagulant, anticardiolipin antibody, and anti-beta-2 glycoprotein I antibodies) were not included in the study. Patients with heparin-induced thrombocytopenia (HIT), idiopathic thrombocytopenia which was thought to be of immune origin, and those with thrombotic thrombocytopenic purpura were excluded from the study.

The detection of SARS-CoV-2 was conducted by real-time reverse transcriptase–polymerase chain reaction method via a nasopharyngeal swab. Patients were divided into two groups as thrombocytopenic (<150 × 109/L) patients on admission or those who developed thrombocytopenia during ICU follow-up (Group 1) and nonthrombocytopenic patients who never had thrombocytopenia during ICU follow-up (Group 2). In the SARS-CoV-2 outbreak, thrombocytopenia was reported to occur in up to 55% of patients and was identified as a significant risk factor for mortality.[11] Thrombocytopenia was accepted as <150 × 109/L in the study by Guan et al. and in many studies in the recent meta-analysis by Lippi et al. Therefore, in our study, the threshold for thrombocytopenia was accepted as <150 × 109/L.[3]

Demographic characteristics, clinical manifestations, treatment, and laboratory findings were extracted medical records from electronic Origo® Hospital Information Management System (Origo, Reykjavík, Iceland). ICU length of stay, hospital length of stay, and death were also recorded. SOFA score and Acute Physiology and Chronic Health Evaluation II (APACHE II) score were calculated within 24 h of ICU admission.

All patients received favipiravir (800 mg per oral (PO) twice daily for 2 days, followed by 600 mg PO twice daily to complete a 5-day course) and subcutaneous prophylactic enoxaparin (0.1 mg/kg/day) during their hospitalization in accordance with the COVID-19 treatment guidelines of the Turkish Ministry of Health. The study was approved by the institutional review board/ethics committee (approval number: 107/2, March 22, 2021). Written informed consent was obtained from the patients when possible; however; surrogate consent was permitted in cases where the patient did not have enough consciousness to make a decision. This manuscript adheres to the STROBE guidelines.

Statistical analysis

IBM SPSS Statistics version 20.0 (IBM Corp., Armonk, NY, USA) was used for statistical analyses. In statistical analysis, categorical variables were given as numbers and percentages, and continuous variables were presented with mean ± standard deviation for descriptive analyses. Chi-square tests were used for the comparison of categorical variables between the groups. The Mann–Whitney U tests were used for comparing the differences in continuous variables between the groups. P < 0.05 was considered statistically significant. Kaplan–Meier analysis and log-rank test were performed to assess ICU and hospital length of stay between thrombocytopenic and nonthrombocytopenic groups.

RESULTS

A total of 181 patients who were over 18 years old were evaluated. One patient with immune thrombocytopenia, two patients with myelodysplastic syndrome, and three patients with positive immune markers were excluded from the study and 175 patients were included in the study. Out of them, 78 (44.6%) were female and 97 (55.4%) were male, with the mean age of 66.2 ± 14.9 years. The mean APACHE II score was 17.8 ± 7.6 and the mean SOFA score was 4.3 ± 2.1 [Table 1].

Table 1

General characteristics of all study groups

Characteristics	n=175
Age (mean±SD)	66.2±14.9
Gender (man), n (%)	97 (55.4)
APACHE II (mean±SD)	17.8±7.6
SOFA (mean±SD)	4.3±2.1
Platelet (×109/L)	240×109 (15–596)
D-dimer (mg/ml) (median, interquartile range 25–75)	1.16 (0.125–31.81)
Presence of thrombocytopenia at admission (%)	32 (18.3)

APACHE II: Acute Physiology and Chronic Health Evaluation II score, SOFA: Sequential organ failure assessment score, SD: Standard deviation

On admission, 32 (18.3%) patients had thrombocytopenia. The overall median platelet count was 240 × 109/L (55–596 × 109/L) and the overall median D-dimer was 1.16 mg/ml (0.125–31.81 mg/ml). The mean platelet count of Group 1 (n = 80) was 105.0 ± 47.5 × 109/L. Group 1 was older (72.9 ± 10.0 vs. 61.4 ± 16.7 years; P = 0.014) and their APACHE II and SOFA scores were higher [19.5 ± 7.8 vs. 16.3 ± 7.2; P = 0.006; Table 2].

Table 2

Comparison of general characteristics, laboratory values, and intensive care unit outcomes of thrombocytopenic (Group 1) and nonthrombocytopenic (Group 2) patients

Characteristics	Group 1	Group 2	P
Age (year)	72.9±10.0	61.4±16.7	0.014*
Gender (male), n (%)	43 (50.0)	54 (50.9)	0.311
APACHE II	19.5±7.8	16.3±7.2	0.006*
SOFA	5.2±2.2	3.5±1.7	0.014*
Labaratory values at the time of ICU admission
Hemoglobin (gr/dl)	12.1±2.4	12.6±2.0	0.046*
Neutrophil count (103/µL)	7.8±5.6	8.2±4.2	0.012*
Lymphocyte count (103/µL)	0.66 (0.41–0.86)	0.88 (0.58–1.24)	0.050**
BUN (mg/dL)	82.7±47.4	51.2±34.2	0.001*
Creatinine (mg/dL)	1.6±1.3	1.0±0.7	0.001*
Procalcitonin (µg/L)	0.7 (0.15–3.6)	0.21 (0.1–0.7)	0.022**
Ferritin (µg/L)	1180±1026	914±726	0.007*
CK (U/L)	538.6±127	206.2±114.7	0.002*
Aptt (sn)	42.9±12.3	35.6±9.0	0.013*
D-dimer (µg/ml)	3.9±2.7	4.75±3.13	0.086
PaO2/FiO2 (mmHg)	123.7±65.7	143.7±72.3	0.086
Anticoagulation use in ICU follow-up (%)	92.5	98.9	0.048*
The prevelance of DIC, n (%)	23 (28.8)	3 (3.2)	<0.001**
Type of oxygen support therapy (%)
Low flow oxygen therapy (nasal cannula and mask)	45.7	54.3	0.26
NIMV	6.3	78.8	0.091
HFNO	70.0	56.8	0.085
IMV	57.5	32.6	0.001
The result of ICU
Length of stay in ICU (day)	11.9±7.2	8.5±5.5	0.001*
Length of stay in hospital (day)	17.3±10.0	14.7±8.4	0.048*
Transfer to inpatient service (%)	32.5	61.1	0.001
Exitus (%)	61.3	31.6	0.001*

*Mean±SD, **Median, interquartile range (25–75). APACHE II: Acute Physiology and Chronic Health Evaluation II score, SOFA: Sequential organ failure assessment score, PaO2/FiO2: Arterial oxygen partial pressure/fractional inspired oxygen, BUN: Blood urea nitrogen, CK: Creatinine kinase, APTT: Activated prothrombin time, ICU: Intensive care unit, DIC: Disseminated intravascular coagulation, NIMV: Noninvasive mechanical ventilation, HFNO: High-flow nasal oxygen therapy, IMV: Invasive mechanical ventilation, SD: Standard deviation

The development of thrombocytopenia in the follow-up caused a 0.4-fold increase in mortality in COVID-19 patients (95% confidence interval: 0.385–0.510). The duration of ICU stay (11.9 ± 7.2 vs. 8.5 ± 5.5 days; P = 0.001) and total hospitalization duration (17.3 ± 10.0 vs. 14.7 ± 8.4 days; P = 0.048) of Group 1 were longer [Figure 1a and b]. The ICU mortality of Group 1 was higher (61.3% vs. 31.6%; P = 0.001) and fewer patients were discharged to the wards from ICU (32.5% vs. 61.1%; P = 0.001). Invasive mechanical ventilation (IMV) was more commonly needed (57.5% vs. 32.6%; P = 0.001), and DIC was more frequently observed (28.5% vs. 3.2%; P = 0.001) in Group 1 [Table 2].

Figure 1

(a) Comparison of the duration of intensive care unit stay of Group 1 (thrombocytopenic) and Group 2 (without thrombocytopenia). (b) Comparison of the duration of the hospital stay of Group 1 (thrombocytopenic) and Group 2 (without thrombocytopenia)

Group 1 had lower hemoglobin level (16.3 ± 7.2 vs. 19.5 ± 7.8 gr/dl; P = 0.006), neutrophil count (7.8 ± 5.6 vs. 8.2 ± 4.2 × 103/μL; P = 0.012), and lymphocyte count (0.66 × 103/μL (0.41–0.86) vs. 0.88 × 103/μL (0.58–1.24), P = 0.001) and higher ferritin level (1180 ± 10.6 vs. 914 ± 726 μg/L; P = 0.007) and procalcitonin (0.7 μg/L [0.15–3.26 μg/L] vs. 0.21 μg/L [0.1–0.7 μg/L]; P = 0.022) level. Group 1 had higher blood urea nitrogen (BUN; 82.7 ± 47.4 vs. 51.2 ± 34.2 mg/dl; P = 0.001), creatinine (1.6 ± 1.3 vs. 1.0 ± 0.7 mg/dl; P = 0.001), creatinine kinase (538.6 ± 127 vs. 206.2 ± 114.7 U/L; P = 0.002), and activated partial thromboplastin time (aPTT) [42.9 ± 12.3 vs. 35.6 ± 9.0 seconds; P = 0.013; Table 2].

DISCUSSION

Thrombocytopenia is common in patients with COVID-19, and it is associated with an increased risk of mortality.[3] In our study, nearly half of the patients had thrombocytopenia (overall, both on admission or during ICU stay at any time) and thrombocytopenic patients had higher mortality; length of ICU stay and hospitalization stay were longer.

Guan et al.[12] reported the incidence of thrombocytopenia (defined by platelet count <150 × 109/L as in our study) in up to 36% of COVID-19 patients. The prevalence of thrombocytopenia in our study was lower than the prevalence previously observed in the literature. The population of this study was older and more severe than the initial studies. The thrombocytopenic patient group had older age (72.9 ± 10.0 vs. 61.4 ± 16.7); thus their bone marrow and thrombocyte series could be more susceptible to be suppressed due to viral infection.

Mounting evidence suggests that lower platelet counts negatively correlate with mortality in COVID-19 patients.[1314] Coagulation abnormalities in COVID-19 disease such as prothrombin time and aPTT prolongation, increase of fibrin degradation products, and severe thrombocytopenia might lead to life-threatening DIC, which necessitates continuous vigilance and prompt intervention.[13] In our study, thrombocytopenic patients had higher APACHE II and SOFA scores, had lower hemoglobin, neutrophil count, and lymphocyte count and higher ferritin and procalcitonin level. The need for IMV, the signs of DIC, the duration of ICU stay and total hospitalization duration, and overall mortality were also higher in thrombocytopenic patients.

Causes of decreased platelet counts in COVID-19 include the endovascular pathophysiology of COVID-19 itself, DIC, antiphospholipid antibody syndrome, immune thrombocytopenic purpura (ITP), hemophagocytic syndrome, HIT, drug-induced myelosuppression, and pseudo-thrombocytopenia.[15] Another important aspect is the thrombocytopenia is whether the emergence of thrombocytopenia result in increased disease severity or the increased severity of the disease decreases the platelets.[16]

Yang et al.[13] investigated the relationship between prognosis of COVID-19 and platelet count and found 1476 thrombocytopenic COVID-19 cases from a single facility, of whom 238 patients died. They also reported that when the platelet count fell below 10 × 103/μL during the course of treatment, the prognosis became particularly severe, whereas if it dropped below 5 × 103/μL, the patient was considered to be in DIC and the mortality seemed inevitable if no intervention was made.

Many drugs used during the treatment of COVID-19 may result in drug-induced decreases in platelet production (myelosuppression). Such circumstances can be easily distinguished by checking the immature platelet fraction (IPF), mean platelet volume (MPV), and platelet distribution width (PDW). In fact, IPF correlates positively with MPV and PDW.[17] Although MPV and PDW are advantageous due to the fact that the data can be collected using standard equipment in all medical institutions, and for all cases, one limitation of MPV and PDW is that they cannot be calculated if the platelet count is significantly reduced.

Antiphospholipid antibodies (anticardiolipin antibody, lupus anticoagulant, and beta-2 glycoprotein I) have long been known to show temporary positivity in viral infections and appear at a high rate in COVID-19 patients.[18] The extent to which antiphospholipid antibodies in COVID-19 are associated with thrombosis warrants further study. ITP has been reported in COVID-19. If a sharp and marked drop in platelet count is present, ITP is possible if DIC is negative.[19] Hemophagocytic lymphohistiocytosis (HLH) has also been reported with cytokine storms; however, in COVID-19 cases, performance of bone marrow examination is presumably difficult, so many cases of HLH may remain undiagnosed.[20]

Treatment of COVID-19 is classified into antiviral treatment, cytokine storm treatment, and thrombosis treatment.[15] Rather than providing uniform treatment, the treatment method most suitable for the severity and stage should be selected for each patient. Patients included in this study received favipiravir as the antiviral treatment, and enoxaparin for thrombosis prophylaxis. Heparin is often used during the treatment of COVID-19, and HIT is often reported.[21] This diagnosis should not be delayed, and proactive measurement of HIT antibodies is needed. In our study, low-molecular weight heparin (enoxaparin) was used for venous thromboembolism prophylaxis in all patients. Patients who were thought to have thrombocytopenia due to enoxaparin in the clinical follow-up were excluded from our study.

Elevated serum ferritin levels, procalcitonin levels, interleukin-6 (IL-6) levels, and CRP levels have been linked with poor prognosis and a higher risk of ARDS and death in COVID-19 patients.[2223] Furthermore, blood hypercoagulability is common among hospitalized COVID-19 patients.[1823] Elevated D-dimer levels are consistently reported, whereas their gradual increase during the disease course is particularly associated with disease worsening. Elevated ferritin and procalcitonin indicate the severity of inflammation, which refers to platelet hyperactivity and consumption. In our study, ferritin level of thrombocytopenic patients was significantly higher, indicating that the disease severity was higher in this group.

In their study comparing COVID-19 patients with thrombocytopenia and nonthrombocytopenia, Zhu et al. found significantly lower white blood cells and higher BUN and SOFA score, and higher creatinine and procalcitonin in the thrombocytopenia group, although not significant.[24] They also reported that older age, lower platelet count, and longer aPTT at admission were determined to be risk factors of 28-day mortality.[24]

Recent reports involving COVID-19 showed platelet hyperactivity, thrombocytopenia as well as higher D-dimer levels, prolonged prothrombin time indicates hypercoagulopathy with increased disease severity.[25] Platelets have important functions not only in hemostasis but also in inflammatory processes through the generation of thrombopoietin, which stimulates IL-6 production and produces pro-inflammatory and procoagulant processes. In recent studies, a higher MPV level was observed in thrombocytopenic patients who also have higher mortality.[26] This can be explained by the fact that large platelets are more active, and rapidly migrate to the inflammatory area; thus they are consumed faster.

However, our study has some limitations. First, the acquisition of data was retrospective for the patients included in the cohort. Second, the reason behind the thrombocytopenia was not clear; however, the aim of this study was to assess the impact of thrombocytopenia upon ICU admission on poor outcomes, regardless of the mechanism.

CONCLUSION

Investigation of laboratory parameters will help physicians to predict outcomes and severity of disease, and to manage potential complications such as DIC. A main strategy for the future of COVID-19 should aim at building predictive models that would combine inflammatory and coagulation markers with clinical features. The presence of thrombocytopenia may reflect the progression of COVID-19 toward an unfavorable outcome.

Research quality and ethics statement

This study was approved by the Institutional Review Board/Ethics Committee at the University of Health Sciences, Dışkapı Yıldırım Beyazıt Research and Education Hospital (Approval # 107/2; Approval date March 22, 2021). The authors followed the applicable EQUATOR Network (http://www.equator-network.org/) guidelines, specifically the STROBE Guidelines, during the conduct of this research project.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.