Dysregulated platelet function in patients with postacute sequelae of COVID-19

Abstract

Background:

Postacute sequelae of COVID-19 (PASC), also referred to as “Long COVID”, sometimes follows COVID-19, a disease caused by SARS-CoV-2. Although SARS-CoV-2 is well known to promote a prothrombotic state, less is known about the thrombosis risk in PASC. Our objective was to evaluate platelet function and thrombotic potential in patients following recovery from SARS-CoV-2, but with clear symptoms of patients with PASC.

Methods:

patients with PASC and matched healthy controls were enrolled in the study on average 15 months after documented SARS-CoV-2 infection. Platelet activation was evaluated by light transmission aggregometry (LTA) and flow cytometry in response to platelet surface receptor agonists. Thrombosis in platelet-deplete plasma was evaluated by Factor Xa activity. A microfluidics system assessed thrombosis in whole blood under shear stress conditions.

Results:

A mild increase in platelet aggregation in patients with PASC through the thromboxane receptor was observed, and platelet activation through the glycoprotein VI (GPVI) receptor was decreased in patients with PASC compared to age- and sex-matched healthy controls. Thrombosis under shear conditions as well as Factor Xa activity were reduced in patients with PASC. Plasma from patients with PASC was an extremely potent activator of washed, healthy platelets – a phenomenon not observed when stimulating healthy platelets after incubation with plasma from healthy individuals.

Conclusions:

patients with PASC show dysregulated responses in platelets and coagulation in plasma, likely caused by a circulating molecule that promotes thrombosis. A hitherto undescribed protective response appears to exist in patients with PASC to counterbalance ongoing thrombosis that is common to SARS-CoV-2 infection.

Background

COVID-19, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), may have residual effects following recovery. One-third of infected patients are asymptomatic, whereas others experience pulmonary and cardiovascular symptoms leading to morbidity and mortality.¹ The risk of both severe and persistent symptoms of SARS-CoV-2 is higher men, in the advanced-age patient population, in patients requiring immunosuppression, or with chronic kidney disease.² The risk of severe and persistent symptoms of SARS-CoV-2 remains despite vaccination.^3,4

Postacute sequelae of COVID-19 (PASC) is a recognized multisystemic condition characterized by persistent symptoms 4 weeks beyond the initial SARS-CoV-2 infection. Though acute respiratory distress syndrome (ARDS) is a feared consequence of severe COVID-19, many acutely ill patients are reported to have thromboembolic complications including deep vein thrombosis (DVT) and pulmonary embolism (PE). The incidence of venous thromboembolic (VTE) complications may be as high as 40%.^5,6

Observational data suggest that patients with PASC develop or have persistent symptoms long after recovery that were initially labeled Long COVID by the Centers for Disease Control (CDC) and the National Institutes of Health (NIH). The National Institute for Health and Care Excellence (NICE) now defines PASC as postinfection symptoms beyond 4–12 weeks.^7-9 Patients blighted by the specter of PASC show structural and functional organ derangements involving the respiratory, cardiovascular, neurological, genitourinary, gastrointestinal, and musculoskeletal systems. PASC symptoms typically include fatigue, dyspnea, chest pain/tightness, headache, cough, difficulty concentrating, and altered sensorium. Poor memory and concentration are the most frequent symptoms reported, closely followed by cognitive impairment, sleep disturbance, and posttraumatic stress disorder (PTSD).^9-13

Elevated prothrombotic markers including D-dimer, C-reactive protein, and neutrophil extracellular trap (NET) formation have been detected in acute SARS-CoV-2 infection. In addition, many patients with PASC complain of dyspnea on exertion, which leads to physical inactivity and may further precipitate thromboembolic complications. Virus-induced endothelial injury of severe COVID-19 patients initiates a state of dysregulated coagulation. Increased blood von Willebrand factor (VWF), likely from both endothelial cell and platelet activation, promotes hypercoagulability and thrombosis in patients with COVID-19.^7,8,14

Platelets have an important function to prevent excessive bleeding (hemostasis) alongside an important role in the host immune response to invading viral and microbial infections, especially SARS-CoV-2.¹⁵ In a large observational study, no additional risk for VTE after hospitalization with acute SARS-CoV-2 infection was noted.¹⁶ Contrasting this, the MICHELLE study revealed a protective effect of an oral Factor Xa inhibitor on discharged patients by protecting them against DVT if they were previously hospitalized, even without an initial thrombotic event.¹⁷ These findings raise the possibility that following SARS-CoV-2 recovery, patients may have ongoing functional impairment in the coagulation cascade or abnormal platelet function. The aim of this study was to evaluate both platelet function and thrombotic potential in whole blood, in platelet-deplete plasma, and in isolated platelets in patients following SARS-CoV-2 but with clear PASC symptoms.

Methods

A total of 26 patients with PASC at a single institution in Northeast Ohio were enrolled in this study on average 501 days following diagnosis of SARS-CoV-2 infection, with persistent symptoms including shortness of breath, fatigue, brain fog, dyspnea on exertion, mood disturbance, weakness, and memory impairment. A total of 19 matched comparators were enrolled from the same geographic location. Patients and healthy subjects were recruited only after informed consent was obtained. All study procedures were conducted in accordance with the Declaration of Helsinki and approved by a local institutional review board (IRB) protocol. Human platelet physiological studies must be conducted within 90 minutes for meaningful data to be obtained. These studies are labor-intensive and, in their entirety, cannot be conducted simultaneously for every patient within 90 minutes. The volume of blood required to conduct studies in whole blood, in platelet-rich plasma (PRP), in washed platelets, and in platelet-deplete plasma is also substantial and exceeds the blood volume permitted to be drawn in the IRB protocol. Therefore, it is impossible to conduct every experiment in every enrolled PASC patient or healthy individual. Rather, the studies were conducted in a manner that was practically required as patients with PASC were identified.

Cross-activation study using flow cytometry

To check whether plasma from patients with PASC would modify platelet activation from healthy donors, we performed cross-activation tests. Platelet-deplete plasma (20 μL) from the patients with PASC (n = 18) as well as age- and sex-matched healthy controls (n = 20) was added to the isolated, washed platelets (80 μL) from healthy controls (each performed in triplicate) and incubated for 20 minutes at room temperature. Platelets were stimulated with agonists: thrombin receptor activator peptide 6 (TRAP-6; 5 μM), thromboxane A₂ (TXA₂) analogue U46619 (1 μM), adenosine diphosphate (ADP; 0.1 μM), or collagen-related peptide (CRP; 0.1 μg/mL) for 15 minutes. Samples were next stained with phycoerythrin (PE)-conjugated anti-CD62P (P-selectin) for 30 minutes at room temperature in darkness to prevent fluorophore bleaching before fixing with 4% paraformaldehyde. Fixed samples were read by flow cytometry (BD Accuri C6 Plus; BD Biosciences, Franklin Lakes, NJ) to acquire 20,000 events and data were analyzed using FlowJo software.

Antibody profiling

Antibody profiling was performed by CDI Labs (Baltimore, MD) using the HuProt v4.0 array. Briefly, sequence-confirmed plasmids, which are used to make GST-purified recombinant proteins in yeast. After purification from yeast, GST-tagged proteins are piezoelectrically printed on glass slides in duplicate, along with control proteins (GST, BSA, histones, IgG, etc.). The slides are barcoded for tracking and archiving. Each microarray batch is routinely evaluated by anti-GST staining to demonstrate quality of expression and printing. The new HuProt v4.0 consists of unique human proteins, isoform variants, and protein fragments of the 19,613 canonical human proteins described in the Human Protein Atlas.^18-21 We focused on protein exclusively expressed in platelets, highly expressed in platelets, and blood proteins known to be involved in the alterations in coagulation as noted.

Content includes major functional classes such as intracellular proteins, membrane proteins, enzymes, secreted proteins, transcription factors, transporters, G protein-coupled receptors (GPCRs), cytokines, immune receptors, immune checkpoints, CD markers, ion channels, cytosolic proteins, and nuclear receptors. To measure antibody binding, slides are overlaid with a 1:100 dilution of serum and stained with antiimmunoglobulin G (IgG) (red) and antiimmunoglobulin M (IgM) (green) secondary antibodies. A heat map was generated to show each antibody concentration, expressed as SD above or below the mean concentration found in healthy individuals.

Light transmission aggregometry (LTA)

Blood from healthy controls and patients with PASC was collected in sodium-citrate tubes after obtaining informed consent. Within 30 minutes to 1 hour of collection, anticoagulated whole blood was centrifuged at 200 g for 15 minutes at room temperature (RT) to obtain platelet-rich plasma (PRP). PRP was collected in 15 mL Falcon tubes and platelet-poor plasma (PPP) was obtained from the remaining sample by recentrifugation at 2500 g for 10 minutes. PPP was used to initiate baseline optical density. Platelet aggregation was tested following stimulation with agonists: ADP (1 μM), TRAP-6 (5 μM), U46619 (5 μM), and CRP (1 μg/mL). Optical density changes were recorded after the addition of the agonists on an aggregometer (CHRONO-LOG 700; Havertown, PA) for 6 minutes while stirring the sample at 1200 rpm at 37°C.

Alpha-granule exocytosis with flow cytometry

Platelets were isolated from whole blood collected from healthy controls and patients with PASC in sodium-citrate tubes. Briefly, whole blood was centrifuged at 200 g for 15 minutes at RT and then PRP was collected in the 15 mL Falcon tubes. The collected PRP was centrifuged again at 200 g for 10 minutes to remove remaining white blood cells and red blood cells. Platelets were washed using Tyrode’s buffer in the 1:1 ratio with the addition of prostaglandin I₂ (PGI₂; 10 nM). Washed platelets were resuspended in 10 mL of Tyrode’s buffer. Washed platelets from patients with PASC and healthy controls were stimulated using different concentrations of agonist: TRAP-6 (1, 5, 10, 20 μM), U46619 (0.1, 0.5, 1.0, 10 μM), ADP (0.01, 0.1, 1.0, 10 μM), and CRP (0.05, 0.1, 0.5, 1.0 μg/mL) for 15 minutes and then incubated with PE-conjugated anti-CD62P for 30 minutes at RT. Samples were fixed with 4% paraformaldehyde. P-selectin expressed on the surface of the platelets signifies an activated state. Samples were read by flow cytometry (BD Accuri C6 Plus) to acquire 20,000 events. Data were analyzed using FlowJo software.

Total Thrombus formation Analysis System (T-TAS01)

The T-TAS01 microfluidic chamber system was used, and the results were compared to healthy controls. T-TAS01 is a microchip flow-chamber system that measures thrombus formation under blood flow conditions over time. Whole blood collected in the sodium-citrate tube was used for the test. Briefly, before measurement, 20 μL of the CaCTI reagent (Corn trypsin inhibitor) was added to 480 μL of whole blood and then this blood was mixed gently with the pipette. Then 450 μL of mixed blood was loaded into the reservoir of the AR chip coated with both collagen and thromboplastin (tissue factor) with a shear rate of 600 s⁻¹. As coagulation and platelet activation progress, thrombosis is noted as pressure rises in the channel following shear rate-mediated activation of blood in the microchip channel. These pressure changes are monitored by a pressure sensor upstream of the chamber and are defined as the occlusion start time (OST) and the occlusion time (OT). The difference between OT and OST was calculated for each subject. Furthermore, to check the prothrombotic activity of the plasma of patients with PASC, 200 μL plasma from patients with PASC and healthy controls was incubated with 800 μL whole blood from healthy controls for 30 minutes. Then this spiked whole blood with plasma was loaded into the reservoir of the AR chip.

Plasma Factor Xa activity

Factor Xa is the activated form of the coagulation factor X. Factor X, a serine endopeptidase, plays an important role in the coagulation pathway. It converts prothrombin into active thrombin by complexing with activated co-factor V in the prothrombinase complex. The Factor Xa Activity Assay Kit from Abcam (ab204711; Cambridge, MA) was used to measure the Factor Xa activity in the plasma of the patients with PASC compared to age- and sex-matched healthy controls. The assay was performed according to the manufacturer’s recommendations.

Statistical analysis

All data were interrogated for normalcy using the Shapiro–Wilk test. Normally distributed continuous variable group differences were assessed by the two-tailed Student’s t-test and skewed data assessed by the Mann–Whitney U-test. Analyses were conducted using GraphPad Prism 7 (GraphPad Software). For Gaussian-distributed data in three or more groups, one-way ANOVA then the Bonferroni multiple comparisons test were used, otherwise the Kruskal–Wallis test followed by the Dunn post-test were used. Significance was accepted as a p-value of less than 0.05.

Results

Baseline study population

A total of 26 patients with PASC (mean age 49.9) were recruited from the pulmonary medicine clinic in Northeast Ohio and compared with 19 healthy volunteers (mean age 41.4). Healthy recruits or patients with PASC were comparable, showing no significant difference between groups with respect to sex, body mass index (BMI), and medical comorbidities; although patients with PASC were more commonly smokers (Table 1). Patients with PASC presented on average 501 days ± 69 days after the last detection of a SARS-CoV-2 amplicon by polymerase chain reaction (PCR). Of the 26 common PASC symptoms in our patient population, most described fatigue (76.9%), brain fog (76.9%), shortness of breath at rest (69.2%), and exertional dyspnea (69.2%) (online Supplemental Table 1).

Table 1.

Demographics for patients with PASC and healthy controls.

Demographics	Healthy (n = 19)		PASC (n = 26)		p-value
	n	%	n	%
Characteristics
Mean age (± SEM)	41.4 ± 2.89		49.9 ± 3.61		0.090
18–27 years	4	21.0	3	11.5	0.30
28–40 years	4	21.0	7	26.9	0.76
41–60 years	9	47.3	5	19.2	0.51
61–82 years	2	10.5	11	42.3	0.62
Men	4	21.0	7	35.0	0.73
Women	15	78.9	19	73.0
Mean BMI (± SEM)	29.5 ± 1.29		30.9 ± 2.17		0.61
Smoker	1	7.7	13	50.0	0.002
Medical history
Hypertension	4	21.0	6	23.0	0.99
Hyperlipidemia	2	10.5	5	19.2	0.68
History of CVD	1	5.2	3	11.5	0.62
Diabetes	2	10.5	1	3.8	0.56
History of pulmonary embolism	0	0.0	1	3.8	0.99
History of thrombosis	1	5.2	0	0.0	0.42
Medications
Platelet inhibitors	1	5.2	6	23.0	0.21
Aspirin	1	5.2	5	19.2
DAPT			1	3.8
Anticoagulants	0	0.0	2	7.6	0.50
Warfarin			1	3.8
Apixaban			1	3.8
Statins	2	15.4	5	19.2	0.68
NSAIDs	6	31.5	4	15.4	0.28
BP medications	4	21.0	1	3.8	0.14

Differences between groups were calculated by Student’s t-test for continuous variables or chi-squared test for dichotomous variables. Bold indicates a statistically significant difference.

BMI, body mass index; BP, blood pressure; CVD, cardiovascular disease; DAPT, dual antiplatelet therapy; NSAID, nonsteroidal antiinflammatory drugs; PASC, postacute sequelae of COVID-19.

Platelet reactivity in patients with PASC

When platelet-deplete plasma from healthy individuals or patients with PASC was added to healthy, twice washed platelets, there was a surprisingly profound activation of platelets following stimulation with all platelet receptor agonists tested, caused by PASC patient plasma incubation (Figure 1). This effect was not observed when platelet-deplete plasma from sex- and age-matched healthy subjects was added to healthy, twice-washed platelets. Given that immunothrombosis is common in COVID-19 and the complement system likely contributes to these responses,^22-24 we heated and inactivated plasma from PASC donors at 55°C and repeated the experiments. We observed that platelets were persistently activated if spiked with room temperature plasma or heat-inactivated plasma prior to agonist stimulation from patients with PASC, suggesting that complement activation is unlikely to be a driver in the augmented platelet response observed (online Supplemental Figure S1).

Figure 1.

Activation of platelets by PASC but not healthy plasma. Isolated, washed platelets were preincubated for 30 minutes with 20 μL of age- and sex-matched healthy plasma or PASC plasma, then stimulated with TRAP-6, 5 μM (PAR1), U46619, 1 μM (thromboxane receptor), ADP, 0.1 μM (P2Y₁₂ receptor), or CRP, 0.1 μg/mL (GPVI receptor) for 15 minutes. Platelet reactivity was assessed by FACS by surface CD62P (P-selectin) using a tagged antibody. Washed platelets were from a healthy 25-year-old man.

Data are represented as mean ± SEM.

Group differences were assessed by the Mann–Whitney U-test: n = 20 for plasma isolated for healthy individuals, n = 18 for plasma isolated for patients with PASC.

ADP, adenosine diphosphate; CRP, collagen-related peptide; FACS, fluorescence-activated cell sorting; PASC, postacute sequelae of COVID-19; TRAP-6, thrombin receptor activator peptide 6.

Platelet aggregation in PRP was evaluated by LTA after stimulating platelets with surface receptor agonists for the P2Y₁₂ receptor (ADP), the thromboxane receptor (U46619), the GPVI receptor (CRP), and PAR1 (TRAP-6). Overall, platelet aggregation by LTA in PASC was similar to healthy individuals except through the thromboxane receptor which demonstrated increased reactivity following stimulation with the agonist U46619 (87% ± 3 for PASC vs 72% ± 4 for healthy subjects; p = 0.03) (Figure 2A-D). To distinguish platelet function that may be altered by mediators in PRP, these studies were repeated in a dose–response manner after isolating and washing platelets twice, then assessing platelet reactivity by α-granule exocytosis by appearance of P-selectin on the surface of the platelet, as reported previously.^25,26 When washed platelets from healthy volunteers and patients with PASC were stimulated with platelet surface receptor agonists, we found similar reactivity to healthy individuals through the following receptors: P2Y₁₂, thromboxane, and PAR1 (Figure 3A-C). Curiously, platelet reactivity by α-granule exocytosis through GPVI was markedly decreased in patients with PASC compared with healthy controls (Figure 3D).

Figure 2.

Platelet activation in PASC by aggregometry. Blood was drawn from healthy individuals or patients with PASC and PRP was isolated prior to stimulation with (A) TRAP-6 (5 μM), (B) U46619 (5 μM), (C) ADP (1.0 μM), or (D) CRP (1.0 μg/mL), and observed until platelet aggregation was noted.

Platelet aggregation was evaluated by LTA and data reported as mean % maximum aggregation ± SEM. A representative aggregometry tracing is shown. Group differences were assessed by the Mann–Whitney U-test: n = 5 for healthy individuals, n = 5 for PASC; or Student’s t-test: n = 5 for healthy individuals, n = 5 for PASC.

ADP, adenosine diphosphate; CRP, collagen-related peptide; LTA, light transmission aggregometry; PASC, postacute sequelae of COVID-19; PRP, platelet-rich plasma; TRAP-6, thrombin receptor activator peptide 6.

Figure 3.

Platelet activation in PASC by flow cytometry. Blood was drawn from healthy individuals or patients with PASC and isolated, washed platelets were stimulated with (A) TRAP-6 (0–20 μM), (B) U46619 (0–10 μM), (C) ADP (0–10 μM), or (D) CRP (0–1.0 μg/mL) for 15 minutes, then reactivity was assessed by FACS by fold-change in surface CD62P (P-selectin) using a tagged antibody.

Data are represented as mean fold-change from baseline (vehicle incubation) ± SEM. The p value is indicated above each experimental group.

Group differences were assessed by the Kruskal–Wallis test: n = 10 for healthy individuals, n = 9 for PASC.

ADP, adenosine diphosphate; CRP, collagen-related peptide; FACS, fluorescence-activated cell sorting; NS, not significant; PASC, postacute sequelae of COVID-19; TRAP-6, thrombin receptor activator peptide 6.

To test the hypothesis that changes in platelet function as observed by aggregometry is augmented in PASC following platelet thromboxane receptor stimulation (Figure 2B) or α-granule exocytosis (decreased in PASC following platelet GPVI stimulation; Figure 3D) are a consequence of changes in platelet surface receptor expression, we performed flow cytometry using receptors tagged with a fluorophore, ultimately finding no difference (online Supplemental Figure S2).

Whole blood and platelet-independent coagulation in patients with PASC

Whole blood was drawn from patients who were healthy or with PASC, then passed through a 80 μm channel in a microfluidics system at 600 dynes (s⁻¹) shear stress. This assay was specifically chosen because thrombosis in most patients with SARS-CoV-2 occurs in the venous vasculature, and the assay utilizes tissue factor and collagen to initiate thrombosis, which simulates endothelial damage common to SARS-CoV-2. Overall, the time to pressure increase in the micro-channel (kPa), which is a signature of ex vivo blood clotting as coagulated blood diminishes flow, was prolonged in patients with PASC compared with healthy conditions (OT – OST, 94 ± 8 vs 73 ± 8 seconds, respectively; p = 0.02) (Figure 4A-B). This suggests whole blood takes longer to clot in patients with PASC under shear stress conditions. To test whether a mediator in plasma is responsible for the difference in coagulation observed in PASC, whole blood from a healthy individual was spiked with platelet-deplete plasma from a PASC patient and compared to spiking whole blood with platelet-deplete plasma from a healthy individual. This showed no difference between OT and OST (62 ± 4 healthy plasma vs 66 ± 8 PASC plasma, n = 5 in each group; p = 0.74). We next assessed Factor Xa activity in platelet-deplete plasma, which represents the point of convergence between the extrinsic and the intrinsic arms of the coagulation cascade. We observed Factor Xa activity was reduced in PASC versus healthy individuals (23.6 ± 5 vs 50 ± 8 ng/100 μL, respectively; p = 0.037) (Figure 4C). Complement, which can increase platelet reactivity, is well known to be denatured at 55°C, yet antibodies targeted against platelet proteins or the coagulation cascade may retain some function.^27-29 We therefore performed a limited antibody array using serum from patients with PASC and comparing this to healthy individuals for antibodies targeted against platelet proteins that could alter the platelet function of antibodies targeted against circulating Factor II, Factor II receptor, or Factor X. We did not find a significant difference in either IgM or IgG against those protein targets comparing patients with PASC to healthy controls (online Supplemental Figure S3).

Figure 4.

Whole blood and plasma coagulation in PASC by microfluidics and Factor Xa activity. (A) Blood was isolated and passed through a microfluidics chip using the Total Thrombus formation Analysis System (T-TAS01) at roughly venous shear stress (600 s⁻¹). With clot generation, pressure in the microfluidics chamber (kPa) gradually increases until cessation of blood flow. Data are represented as mean ± SEM. (B) The time from the start of occlusion to cessation of blood was calculated in seconds as mean ± SEM. Group differences were determined by Student’s t-test for patients with PASC (n = 10) or healthy volunteers (n = 12). (C) Citrate plasma devoid of platelets was evaluated for Factor Xa activity using a chromogenic substrate in patients with PASC (n = 7) or healthy volunteers (n = 10). Data are represented as mean ± SEM. Group differences were assessed by Student’s t-test (n = 12) for healthy individuals and for PASC (n = 10). PASC, postacute sequelae of COVID-19.

These experiments suggest there is a change in PASC plasma with dysregulated coagulation compared with healthy conditions that does not involve complement or circulating antibodies. Overall, the observation that washed platelets from patients with PASC are relatively unchanged or less reactive than healthy platelets, and that platelet-deplete plasma in PASC is less likely to clot, suggests a potential adaptive response in the host long after COVID-19 infection, possibly to prevent thrombosis that is common in patients with acute COVID-19 infection.

Discussion

SARS-CoV-2 is well known to precipitate thrombotic events, especially DVT and PE.^30,31 Acute SARS-CoV-2 infection has also been shown to coincide with activated platelets by multiple independent groups.^32-35 Our primary objective was to assess platelet function and thrombosis in patients with PASC given that rheologic manifestations of this syndrome are an under-developed area of investigation. Our major finding in patients more than 1 year following acute SARS-CoV-2 and with clear symptoms of PASC is dysregulated platelet function, with slight platelet activation through the thromboxane receptor, and attenuated activation through the GPVI receptor. We did not observe a change in platelet receptor signaling through PAR1 or P2Y₁₂ in PASC using two well-validated techniques to interrogate platelet reactivity.

This study may be the first to report a functional defect in the coagulation cascade given our finding of reduced activated Factor Xa in platelet-deplete plasma in patients with PASC.³³ Curiously, using radioligand binding technology, or a similar chromogenic substrate for Factor Xa activation that we employ, others have reported that platelets and platelet-derived microparticles may be a source of Factor Xa activity in the circulation.^36-38 Moreover, a putative signaling mechanism between the platelet GPVI receptor and Factor Xa was previously reported.³⁹ In addition, the profoundly under-reactive GPVI receptor identified by flow cytometry for α-granule exocytosis using CRP as the agonist in our study complements the observation of prolonged time to thrombosis in human blood ex vivo under conditions of venous shear stress since collagen is present in the microfluidics channel which activates the GPVI receptor.

The most significant finding in our study is that platelet-deplete plasma in patients with PASC ‘primes’ and markedly activates healthy, washed platelets through every receptor signaling pathway evaluated. We did not see this when platelet-deplete plasma from healthy controls was incubated with washed platelets before agonist stimulation. The hyperreactive platelet effect caused by PASC plasma persisted even after heating plasma to 55°C, suggesting the complement cascade or circulating antibodies are likely not responsible for augmented platelet activation. Using an antibody array, we could not find antibodies directed against platelet protein targets of coagulation cascade proteins in the blood that could explain the observation in patients with PASC. Therefore, we offer the suggestion that a putative prothrombotic factor exists in the blood of patients with PASC, and its presence leads to an adaptive response in the host in which platelet reactivity and thrombosis are paradoxically tipped in the opposite direction, presumably as a protective host response against thrombosis.

Our study in patients with PASC differs from a recent letter by Martins-Gonçalves et al. in patients reporting at least one persistent respiratory symptom after acute hospitalization for COVID-19 and with persistent platelet hyper-reactivity.⁴⁰ Important differences in the enrolled subjects were noted. Firstly, though the average age of patients enrolled by Martins-Gonçalves et al. was similar to our cohort, the represented majority were men, whereas 73% of our patients and 79% of our healthy controls were women. Platelet reactivity has been shown to differ between men and women in health and in cardiovascular disorders.^26,36,41 Secondly, the control group employed by Martins-Gonçalves et al. tested negative for SARS-CoV-2 weekly for almost 2 years prior to enrollment, whereas several of our control comparators reported SARS-CoV-2 infection in the preceding 2 years but without residual symptoms at the time of enrollment, and this was the reason why they were selected as controls. Lastly, Martins-Gonçalves et al. evaluated platelet reactivity in the convalescent setting 1–4 months after acute SARS-CoV-2 infection in 2021 whereas our patients with PASC on average were evaluated > 16 months after the last documented acute SARS-CoV-2 infection in 2023. Clinical outcomes and symptoms following SARS-CoV-2 infection appear to vary depending on vaccination status and the viral strain of the infected individual.⁴² This means that interpretation of data gained from patients evaluated during different waves of the COVID-19 pandemic require careful attention and appropriate interpretation. Given that SARS-CoV-2 was demonstrated to induce morphological changes in platelets and promote apoptosis after viral internalization in an ACE2-dependent and ACE2-independent receptor internalization pathway,¹⁵ platelet reprogramming in patients with PASC may occur.

In a prior observational study, we were the first to produce data during the COVID-19 pandemic that the antiplatelet drug aspirin does not decrease mortality in patients infected with COVID-19.⁴³ This observation was later supported by high quality randomized controlled trials that revealed aspirin or clopidogrel minimally impact patient outcomes following SARS-CoV-2 infection.^44,45 Unlike COVID-19, the current study suggests differences in platelet activation through the thromboxane receptor and the glycoprotein VI receptor. Whether those differences translate to changes in patient outcomes in patients with PASC would be speculative and would require the rigor of a prospective, double-blinded, randomized controlled study.

Conclusion

In summary, patients who have long recovered from acute SARS-CoV-2 but with persistent symptoms under the constellation of PASC show evidence of dysregulated platelet reactivity, alterations in normal coagulation of platelet-deplete plasma, and diminished clotting in whole blood ex vivo under mechanical shear stress conditions. These events are likely a consequence of adapting to a platelet-activating factor in the plasma of patients with PASC. The stimulus behind the exhausted platelet phenomenon, particularly through GPVI, in patients with PASC, should be investigated and identification of the putative prothrombotic mediator should be prioritized given the implications for patients with PASC being considered as donors of blood and plasma products in future.