Early SARS-CoV-2 infection: Platelet-neutrophil complexes and platelet function

Background Conflicting results have been reported on platelet activity ex vivo and responsiveness in vitro among patients with COVID-19 with or without thromboembolic complications. Objectives To assess platelet reactivity in patients with moderate disease at early stages of COVID-19. Methods We performed a prospective, descriptive analysis of 100 consecutive patients presenting with suspected SARS-CoV-2 infection at University Medical Center Freiburg during the first or second wave of the pandemic. Following polymerase chain reaction testing and compliance with study inclusion criteria, 20 SARS-CoV-2–positive and 55 SARS-CoV-2–negative patients (serving as patient controls) were enrolled. In addition, 15 healthy subjects were included. Platelet reactivity was assessed using whole-blood impedance aggregometry and flow cytometry in response to various agonists. Results Platelet aggregation was significantly impaired in the patients with COVID-19 compared with that in the patient controls or healthy subjects. The reduced platelet responsiveness in the patients with COVID-19 was associated with impaired activation of GPIIb/IIIa (αIIbβ3). In contrast, low expression of P-selectin at baseline and intact secretion upon stimulation in vitro suggest that no preactivation in vivo, leading to “exhausted” platelets, had occurred. The proportion of circulating platelet-neutrophil complexes was significantly higher in the patients with COVID-19 (mean ± SD, 41% ± 13%) than in the patient controls (18% ± 7%; 95% CI, 11.1-34.1; P = .0002) or healthy subjects (17% ± 4%; 95% CI, 13.8-33.8; P < .0001). An analysis of neutrophil adhesion receptors revealed upregulation of CD11b (α-subunit of αMβ2) and CD66b (CEACAM8) but not of CD162 (PSGL-1) in the patients with COVID-19. Conclusion Despite reduced platelet responsiveness, platelet-neutrophil complexes are increased at early stages of moderate disease. Thus, this cellular interaction may occur during COVID-19 without preceding platelet activation.


| I N T R O D U C T I O N
COVID-19 is a systemic but predominantly respiratory disease caused by SARS-CoV-2. The first case report from Wuhan indicated an association between pulmonary embolism and SARS-CoV-2 infection [1]. Since then, numerous trials have reported a high rate of venous and arterial thromboembolic events in patients with COVID-19, especially severe courses [2][3][4][5][6]. The underlying mechanisms leading to a prothrombotic state and severe COVID-19 are incompletely understood. It has been discussed that COVID-19-associated coagulopathy differs from other forms of infection-induced hemostatic changes resulting from the release of proinflammatory cytokines and a dysfunctional endothelium [7][8][9].
In general, inflammation and activation of coagulation represent concurrent responses of the host organism to contain and defend against invading pathogens. These multifaceted processes are referred to as "immunothrombosis" or "thromboinflammation" [10,11]. Neutrophils are essential components of innate immunity and are in the frontline of defense against infection, and formation of neutrophil extracellular traps (NETs) has been described to contribute to COVID-19 pathogenesis [12,13]. Moreover, platelets, being part of the innate immune system, can play an important role in viral infections [14,15]. Platelets are activated by NET formation (NETosis) [16] and express a variety of receptors, such as toll-like receptors, that detect viral pathogen-associated molecular patterns [17,18].
Stimulation of these receptors can also result in platelet activation, granular secretion, and aggregation.
We here report that activation of neutrophils and formation of platelet-neutrophil complexes (PNCs) is a feature of the early phase in SARS-CoV-2 infection, whereas platelet reactivity can be reduced at this stage of the disease.

Essentials
• Both hyperreactive and hyporesponsive platelets have been described in patients with COVID-19.
• We studied platelet and neutrophil (re)activity during early SARS-CoV-2 infection.
• The number of platelet-neutrophil complexes was increased despite the absence of activated platelets.
Patients admitted to the Department of Emergency Medicine of University Medical Center Freiburg because of suspected or proven infection with SARS-CoV-2 were eligible for this study. The decision to perform a polymerase chain reaction (PCR) test for SARS-CoV-2 in deep oropharyngeal or nasopharyngeal swabs was made independently of study inclusion by treating physicians. Patients who were or had been tested positive for SARS-CoV-2 within the last 2 weeks were allocated to the positive group; negatively tested patients served as the patient control group. Healthy volunteers (recruited from medical staff and students) served as an additional control group.
Upon informed, written consent by patients and volunteers, hirudinized whole blood (S-Monovette Hirudin 1.6 mL; Sarstedt) was collected on the day of inclusion to perform whole-blood impedance aggregometry (WBIA; Multiplate, Roche Diagnostics, Roche) and flow cytometric analyses. Laboratory parameters were determined by the Freiburg central laboratory (Table) Table).

| Flow cytometry
Flow cytometry was used to determine surface activation markers of platelets (P-selectin, activated GPIIb/IIIa) and of neutrophils (CD11b, CD162, CD66b), and PNCs. Hirudinized whole blood was diluted in a ratio of 1:6 with Dulbecco phosphate-buffered saline (DPBS) with     we used the chi-square test. A 2-tailed P value of <.05 was considered statistically significant. The 95% CI is provided for differences in means among the groups.

| Characteristics of the study population
The SARS-CoV-2-positive and SARS-CoV-2-negative patients displayed a similar sex ratio or demographic and clinical profile (Table).
Likewise, the fibrinogen (Clauss) and VWF plasma levels did not differ between the SARS-CoV-2-positive and SARS-CoV-2-negative patients but were significantly higher in both the patient groups than in the healthy controls (P < .0001). In contrast, the mean platelet count was lower in the SARS-CoV-2-positive patients than in the SARS-CoV-2negative patients (P = .002) or healthy subjects (P = .004). The leukocyte count in the SARS-CoV-2-infected individuals was decreased compared with that in both the control groups (P < .0001 and P = .04, respectively).
The majority of the patients had a mild-to-moderate disease course.

| Platelet α-granule secretion is not affected in SARS-CoV-2-positive patients
Next, we examined whether reduced platelet aggregation responsiveness upon stimulation in vitro results from circulating "exhausted" platelets because of increased activation in vivo. Such a phenomenon has been described in a variety of clinical conditions [29][30][31] and also in patients with COVID-19 [32]. We measured the surface expression of P-selectin, an established marker of platelet activation and an indicator of α-granule secretion. In fact, we did not detect platelet preactivation because the percentage of circulating P-selectin-positive platelets was low at baseline in the SARS-CoV-2-positive patients (5.7% ± 3.6%) and similar compared with that in the SARS-CoV-2-negative patients (4.6% ± 1.9%) and healthy subjects (3.7% ± Other than SARS-CoV-2; non-ICU inpatients: patients receiving non-ICU ward care; patients with infectious disease: respiratory infection, pneumonia, pleurisy, influenza, tonsillitis, peritonsillitis, esophagitis, cholangitis, neutrocytopenic fever, urinary tract infection, infected lymphocele, erysipelas, viral meningitis, cholangiosepsis, urosepsis, septic shock; patients with heart disease: cardiac decompensation, hydropic decompensation, heart failure with preserved ejection fraction, terminal heart failure; patients with cancer: lung cancer, adrenocortical carcinoma, cancer of unknown primary origin; patients with other disorders: gait disorder, chest pain, hemolytic anemia, headache, migraine attack, pulmonary hypertension, chronic obstructive lung disease, dyspnea, pulmonary sarcoidosis. RIEDER ET AL.   Taken together, these data indicate that neutrophil reactivity is increased during early SARS-CoV-2 infection. F I G U R E 2 Agonist-induced platelet aggregation is reduced in SARS-CoV-2-positive patients. Platelet aggregation was analyzed using whole blood impedance aggregometry (Multiplate) using hirudinized whole blood (1:2 diluted) from SARS-CoV-2-positive patients (dark blue circles, n = 20), SARS-CoV-2-negative patients (light blue squares, n = 55), and healthy volunteers (grey triangles, n = 15). Platelet aggregation was induced by stimulation with thrombin receptor-activating peptide 6 (32 μM), adenosine diphosphate (6.5 μM), or arachidonic acid (0.5 mM) and expressed as area under the curve in arbitrary units. The shaded area represents the reference range, as provided by the manufacturer. Data presented are individual values and mean ± SD of each group. Data were analyzed using 1-way analysis of variance with the Bonferroni multiple comparisons test or Welch analysis of variance with the Games-Howell multiple comparisons test. *P < .05; **P < .01; ***P < .001; ****P < .0001. AA, arachidonic acid; ADP, adenosine diphosphate; AUC, area under the curve; TRAP-6, thrombin receptor-activating peptide 6 RIEDER ET AL. The "classical" mechanism by which PNCs are generated is thought to be exclusively initiated by activated platelets [42], and platelet surface expression of P-selectin is a major driver of PNC formation [43]. In line with this, most of the aforementioned studies reported increased numbers of circulating PNCs in parallel with upregulated platelet surface expression of P-selectin. However, the P-selectin expression at the baseline was negligibly low in the SARS-CoV-2-positive patients in the present study (Figure 3), and this is an unlikely explanation for the observed increased number of PNCs.
We, therefore, investigated whether the upregulation of neutrophil adhesion molecules, such as integrin receptor Mac-1 (α M β 2 ; CD11b/CD18), PSGL-1 (CD162), or CAECAM8 (CD66b), could have mediated PNC formation in response to SARS-CoV-2 infection at early stages of mild-to-moderate COVID-19. Indeed, as depicted in Figure 6, the surface expression of CD11b and CD66b (but not of CD162) was significantly increased in the SARS-CoV-2-positive patients at the baseline in comparison with that in the healthy volunteers. In accordance with our data, recent reports have described upregulation of CD11b [38] or CD66b [39]  in comparison with that in the patient controls and healthy volunteers ( Figure 5). This observation requires distinct consideration. One could assume that this ADP-induced effect might have resulted from concurrent platelet activation, leading to a higher proportion of PNCs.
However, ADP is a weak platelet agonist and induced P-selectin expression on platelets, without significant differences among all the groups ( Figure 3). Moreover, such an explanation suggesting a platelet-induced effect would not be consistent with the findings obtained upon exposure to TRAP-6. On the other hand, in comparison with TRAP-6, incubation with ADP resulted in a higher percentage of platelets with GPIIb/IIIa activation ( Figure 4). Constantinescu-Bercu et al. [45] recently showed that activated GPIIb/IIIa binds to SLC44A2 on neutrophils, which could contribute to PNC formation.
However, GPIIb/IIIa activation upon stimulation with ADP was not increased in the patients with COVID-19 compared with that in the controls ( Figure 4). In addition, as discussed below, the low platelet aggregation responsiveness in patients with mild-to-moderate COVID-19 makes it rather unlikely that the effect of ADP on PNCs is mediated by platelet activation. A more conceivable explanation would be direct stimulation of neutrophils by ADP via purinergic receptors. Such a contention is supported by studies demonstrating the expression of a P2Y12 receptor on leukocytes [46], specifically on eosinophils [47].
Altogether, our findings suggest that PNCs at early stages of SARS-CoV-2 infection are initiated by activated neutrophils. Such a contention underlines the role of neutrophils as "frontline defenders" against invading pathogens. However, the "traditional" concept regarding PNC formation suggests that activated platelets interact with neutrophils, whereby platelet activation is considered an essential requirement in this cellular interplay. Along with our study results, we now propose that increased platelet activity is not a mandatory prerequisite for the formation of circulating PNCs during COVID-19.
This conclusion is based on several observations reported here.
First, we observed a reduced platelet aggregation response in patients with mild-to-moderate COVID-19 ( Figure 2). This is consistent with previous studies that performed WBIA to test platelet function. Bertolin et al. [23] demonstrated lower platelet reactivity using WBIA in hospitalized patients with nonsevere COVID-19.
However, reduced aggregation responses were not limited to a nonsevere clinical course of COVID-19. Heinz et al. [24] also observed normal or reduced aggregation responses using WBIA in ICU patients with COVID-19. A longitudinal assessment by another group revealed that platelet aggregation remained reduced even after 14 days of admission to the ICU [25]. Of note, impaired platelet aggregation is not a unique characteristic of SARS-CoV-2 infection because reduced aggregation responses assessed using WBIA have previously been described with other viral infections [48][49][50] and sepsis [51]. This observation is corroborated by our finding that aggregation parameters below the reference range occurred in 40% of the SARS-CoV- . This is in accordance with recent findings by Weiss et al. [52], suggesting uncoupling between α-granule secretion and GPIIb/ IIIa activation in COVID-19. Impaired GPIIb/IIIa activation in patients with COVID-19 upon agonist stimulation in vitro has been described before [21,38,41] and was associated with a fatal outcome [27].
However, the underlying mechanism is still unclear, and the present study cannot provide an explanation for this observation.
Our study has the following limitations. The number of patients with COVID-19 was small, and no control for confounding was undertaken. This is primarily because of the prospective study design enrolling all patients with suspected SARS-CoV-2 infection prior to the PCR test result. Unexpectedly, only approximately one-third of patients with COVID-19-like symptoms turned out to be SARS-CoV-2 positive upon PCR testing or were tested positive before inclusion.
We cannot exclude the fact that the timing of the positive test result may have affected the outcome. In addition, we performed flow cytometry analyses only on a subset of patients because the study was F I G U R E 6 Neutrophil activation is elevated in SARS-CoV-2-positive patients. Surface expression of neutrophil activation markers CD11b (left panels), CD162 (middle panels), and CD66b (right panels) was assessed in hirudinized whole blood (1:6 diluted) from SARS-CoV-2-positive patients (dark blue circles, n = 14), SARS-CoV-2-negative patients (light blue squares, n = 7), and healthy volunteers (grey triangles, n = 7) after incubation with phosphate-buffered saline (baseline; upper panels) or phorbol 12-myristate 13-acetate (PMA, 100 nM; lower panels) for 15 minutes using flow cytometry and quantified based on geometric mean fluorescent intensity. Data presented are individual values and mean ± SD (CD11b PMA, CD162 baseline and PMA, and CD66b baseline and PMA) or median with interquartile range (CD11b baseline) of each group. Data were analyzed using 1-way analysis of variance with the Bonferroni multiple comparisons test or Welch analysis of variance with the Games-Howell multiple comparisons test or Kruskal-Wallis test. *P < .05; **P < .01.
initially designed to examine only platelet aggregation. Because of lacking information on race or ethnicity, we could not control for sociocultural influences. Another limitation is that samples were collected once at the time of admission, which does not allow longitudinal analyses.
In summary, in patients at early stages of COVID-19, the proportion of circulating PNCs is increased. These heterotypic cellular conjugates can be associated with upregulation of neutrophil adhesion molecules and elevated responsiveness of neutrophils upon stimulation in vitro. In contrast, increased activation of circulating platelets or increased responsiveness in vitro is not detectable at this disease stage. We, therefore, conclude that primary platelet activation is not a mandatory condition for the formation of PNCs during COVID-19.