Low-Grade Inflammation in Long COVID Syndrome Sustains a Persistent Platelet Activation Associated With Lung Impairment

Abstract

In the present study, we provide evidence on the potential mechanisms involved in the residual pulmonary impairment described in long COVID syndrome. Data highlight that lung damage is significantly associated with a proinflammatory platelet phenotype, characterized mainly by the formation of platelet-leukocyte aggregates. In ex vivo experiments, long COVID plasma reproduces the platelet activation observed in vivo and highlights low-grade inflammation as a potential underpinning mechanism, exploiting a synergistic activity between C-reactive protein and subthreshold concentrations of interleukin-6. The platelet-activated phenotype is blunted by anti-inflammatory and antiplatelet drugs, suggesting a potential therapeutic option in this clinical setting.

Keywords: antiplatelet drugs; inflammation; long COVID; platelet-leukocyte aggregates; tissue factor.

Graphical abstract

Figure 1

Circulating Levels of CRP, IL-6, D-Dimer, and Fibrinogen Concentrations of (A) CRP, (B) IL-6, (C) D-dimer, and (D) fibrinogen measured in plasma of long COVID, COVID-recovered, and acute COVID-19 patients. Reference range of healthy subjects is shown as a dotted line. Data are reported as individual and mean ± SD. ∗∗P < 0.01; ∗∗∗P < 0.001. Data were compared by Student’s t-test or Wilcoxon test, as appropriate. CRP = C-reactive protein; IL = interleukin; ns = not significant.

Figure 2

Assessment of Platelet Activation and Association Between Residual Parenchymal Damage at CT Scan With Platelet-Leukocyte Aggregate Formation (A) Platelet-associated P-selectin expression as well as (B) PGA and (C) PMA aggregate formation, were evaluated by whole-blood flow cytometry in long COVID (teal dots) in COVID-recovered patients (6-month follow-up; light blue dots) and in the control group (HSs; green dots) and compared with those measured in acute phase patients, indicated by the orange line. Data were analyzed by Student’s t-test or the Wilcoxon test, as appropriate. Correlations between CRP levels of >0.3 mg/dL and the percentages of (D) P-selectin^pos platelets, (E) PGA, and (F) PMA are shown. Correlations between the number of (G) PGA and (H) PMA in long COVID patients with the percentage of residual parenchymal damage measured at CT scan are shown and analyzed by the Spearman correlation test. Data are reported as individual and mean value ± SD of positive cells. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001. CT = computed tomography; HS = healthy subject; PGA = platelet-granulocyte aggregate; PLT = platelets; PMA = platelet-monocyte aggregate; pos = positive; Psel = P-selectin; other abbreviations as in Figure 1.

Figure 3

Association Between Pulmonary Function Values and PLA Formation Correlation between the DL_NO, V_cap, DM, measured and percentage of predicted DL_CO—corrected for hemoglobin levels and assessed by pulmonary function tests—and the number of (A-E) PGA and (F-J) PMA aggregates, measured by whole-blood flow cytometry in the subgroup of long COVID patients. DL_CO = carbon monoxide lung diffusion; DL_NO = nitric oxide lung diffusion; DM = membrane diffusion; V_cap = capillary volume; other abbreviations as in Figure 2.

Figure 4

Analysis of Platelet-Associated TF Expression (A) Platelet surface TF expression was evaluated by whole-blood flow cytometry, and (B-D) platelet-associated TG was measured by calibrated automated thrombogram assay in long COVID and COVID-recovered patients (6-month follow-up) and in the control group (HS). (E, F) TF^pos PGA and PMA in long COVID and COVID-recovered patients and HSs. The mean value measured in COVID patients during the acute phase is displayed with an orange line. Data are reported as individual and mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001. (G) Analysis by imaging flow cytometry of TF expression in PLA of long COVID patients. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. ETP = endogenous thrombin potential; TF = tissue factor; TG = thrombin generation; other abbreviations as in Figures 1 and 2.

Figure 5

Platelet Adhesion on a Collagen-Coated Microfluidic Surface Ex vivo platelet adhesion was evaluated perfusing blood in collagen-coated microchannels. (A) Average number and (B) area of aggregates in long COVID patients (n = 14) HSs (n = 14) for shear rates of 300/s and 1,600/s were calculated. (C) Representative images (original magnification: 10×) of platelet adhesion over collagen-coated (100 mg/mL) microchannels at 300/s (top) and 1,600/s (bottom) for a patient (long COVID) and an HS. Flow is from left to right. Images were acquired after 4 minutes of perfusion. ∗∗P < 0.01. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. HS = health subject.

Figure 6

Ex Vivo Effect of Plasma From Long COVID and COVID-Recovered Patients on HS Platelets Plasma-depleted blood from HSs (n = 3) reconstituted with autologous plasma (green dots) and plasma from long COVID patients (blue dots; n = 9) or from COVID-recovered subjects (light blue dots; n = 9). Percentages of (A) P-selectin^pos platelets, (B) PGA, and (C) PMA are shown. Data are reported as individual and mean value ± SD. ∗∗P < 0.01; ∗∗∗P < 0.001. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. Abbreviations as in Figure 2.

Figure 7

Ex Vivo Effect of CRP and IL-6 on HS Platelet Activation Whole blood from HSs was incubated with CRP (10-100 μg/mL), IL-6 (1-100 pg/mL), or IL-6 (1-100 pg/mL) + CRP (50 μg/mL). Preincubation of platelets with FcγRi (200 ng/mL; 10 minutes) before CRP (50 μg/mL) stimulation confirmed the specificity of the effect. (A, D) P-selectin^pos platelets, (B, E) PGA, and (C, F) PMA were analyzed by flow cytometry, and the percentages of positive cells are reported. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.001. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. FcγRi = Fcγ-receptor inhibitor; other abbreviations as in Figures 1 and 2.

Figure 8

Ex Vivo Effect of Fcγ-Receptor Inhibitor and Tocilizumab on Platelet Activation Induced by Long COVID and COVID-Recovered Patients’ Plasma Whole blood from healthy subjects, preincubated with FcγRi (200 ng/mL, 10 minutes) or tocilizumab (300 μg/mL), was plasma depleted and reconstituted with (A to C) long COVID (n = 9) or (D-F) COVID-recovered patients’ plasma. Percentages of (A, D) P-selectin^pos platelets, (B, E) PGA, and (C, F) PMA are shown. Data are reported as individual and mean ± SD. The mean values measured in the presence of autologous plasma are indicated by the dotted green line. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. Abbreviations as in Figures 1, 2, and 7.

Figure 9

Ex Vivo Effect of Aspirin and P2Y₁₂ Inhibitor on Platelet Activation Induced by Long COVID and COVID-Recovered Patient Plasma Blood from HSs (n = 6), preincubated or not with aspirin (8 μmol/L) or with P2Y₁₂ inhibitor (AR-C69931MX; 1 μmol/L), was plasma-depleted and reconstituted with autologous plasma (green dots) or with plasma from long COVID patients (n = 6-12; blue dots). Percentages of (A) P-selectin^pos platelets as well as (B) PGA and (C) PMA formation were assessed by flow cytometry. Data are reported as mean value ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001. Data were compared by Student’s t-test or the Wilcoxon test, as appropriate. ASA = acetylsalicylic acid; other abbreviations as in Figure 2.