Impaired Secondary Platelet Response in Chronic Kidney Disease as a Consequence of Prior Platelet Activation

Abstract

Patients with chronic kidney disease (CKD) are at increased risk of thrombotic and hemorrhagic complications. Findings on platelet defects in CKD are conflicting. Therefore, we examined platelet function in CKD stage 3 to 5 dialysis patients without antithrombotic therapy, in CKD5D/hemodialysis patients on acetylsalicylic acid (ASA) as well as in a CKD mouse model. Patients with advanced CKD without antithrombotic therapy show platelet preactivation with a partial secondary platelet dysfunction mainly upon collagen/GPVI stimulation. Platelets from hemodialysis patients on ASA showed a less severe CKD-associated secondary platelet dysfunction compared with those not taking ASA, with comparable observations in CKD mice on ASA vs vehicle.

Keywords: acetylsalicylic acid; bleeding; chronic kidney disease; platelets; thrombosis.

Graphical abstract

Figure 1

Reduced CRP-XL–Induced Integrin α_IIbβ₃ Activation and P-Selectin Expression in Platelets of CKD5D Patients Isolated platelets (50 × 10⁶/mL) were left unstimulated or activated with collagen-related peptide, crosslinked (CRP-XL) for 15 minutes. Integrin α_IIbβ₃ activation (A), P-selectin expression (B), and CD63 expression (C) were assessed using the FITC-conjugated PAC-1 antibody, a PE (phycoerythrin)-conjugated anti–P-selectin antibody or an APC (allophycocyanin)-conjugated anti-CD63 antibody, respectively. Values are median with 25th-75th percentiles. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001 (Kruskal Wallis test with Dunn’s post hoc test). Healthy control subjects n = 16-18, chronic kidney disease stage 3 (CKD3) n = 20, CKD4 n = 15, chronic kidney disease stage 5 dialysis (CKD5D) n = 5, and CKD5D acetylsalicylic acid (ASA) n = 5.

Figure 2

Reduced Ex Vivo Platelet Adhesion and Aggregation Under Flow With Increasing CKD Progression Citrate anticoagulated whole blood of healthy control subjects or CKD patients was recalcified in the presence of PPACK and perfused over a collagen type I surface at 1,000 s⁻¹ for 4 minutes. Integrin activation was monitored by determining the amount of bound fluorescent fibrinogen (Fg) to the thrombi. P-selectin expression and phosphatidylserine (PS) exposure were measured by labeling the formed thrombi with an antibody directed against P-selectin and annexin A5, respectively. (A) Representative images of adhesion and aggregation, P-selectin expression, fibrinogen binding and PS exposure. Scale bar = 20 μm. (B) Quantification of the surface area covered by adhered platelets, aggregated platelets, P-selectin–positive platelets, bound fibrinogen, and PS-positive platelets. Values are median with 25th-75th percentiles. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001 (PS exposure, CKD5D vs healthy control subjects P = 0.056): Kruskal Wallis test with Dunn’s post hoc test for comparison vs healthy control group. ^#P < 0.05; ^##P < 0.01; ^###P < 0.001 Mann-Whitney U test for comparison of CKD5D vs CKD5D ASA. Healthy control subjects n = 21, CKD3 n = 24, CKD4 n = 17, CKD5D n = 7, CKD5D ASA n = 11, cardiovascular disease (CVD) n = 5 and CVD ASA n = 10. Abbreviations as in Figure 1.

Figure 3

Impaired Platelet-Mediated Fibrin Formation Under Flow in CKD5 Patients on Hemodialysis Using a co-perfusion system, citrate anticoagulated blood of healthy control subjects or CKD patients was recalcified and subsequently perfused over a collagen/tissue factor (TF) surface for 10 minutes at 1,000 s⁻¹ allowing the formation of thrombi and fibrin. (A) Representative fluorescence images of thrombus (DiOC6-labeled platelets) and fibrin formation after 10 minutes of blood perfusion. Scale bar = 20 μm. (B and C) Quantification of the platelet (B) and fibrin (C) surface area coverage after 10 minutes of perfusion. Values are median with 25th-75th percentiles. ∗P < 0.05; ∗∗P < 0.01 (fibrin, CKD4 vs healthy control subjects P = 0.074) Kruskal Wallis test with Dunn’s post hoc test for comparison vs healthy control group. ^#P < 0.05 (Mann-Whitney U test for comparison of CKD5D vs CKD5D ASA). Healthy control subjects n = 20, CKD3 n = 21, CKD4 n = 17, CKD5D n = 7, CKD5D ASA n = 9, CVD n = 4, and CVD ASA n = 10. HD = hemodialysis; SAC = surface area coverage; other abbreviations as in Figures 1 and 2.

Figure 4

Short-Term Exposure to Uremic Toxins Causes an Impairment of Platelet Adhesion and Aggregation Under Flow Citrate anticoagulated whole blood of healthy control subjects was treated with a mixture of phenylacetic acid, indoxyl sulphate, hippuric acid, kynurenic acid, p-cresyl sulphate, methylguanidine, and guanidinosuccinic acid, at concentrations reflecting advanced CKD, or alternatively, with vehicle control. After a 5-minute incubation at room temperature, the blood was recalcified in the presence of PPACK and perfused over a collagen type I surface at 1,000 s⁻¹ for 4 minutes. Integrin activation was assessed by the amount of bound fluorescent fibrinogen. After blood perfusion, P-selectin expression and PS exposure were detected by perfusing buffer with an anti–P-selectin antibody and annexin A5, respectively. Quantification of the surface area covered by adhered platelets, aggregated platelets, P-selectin–positive platelets, bound fibrinogen, and PS-positive platelets. ∗∗P < 0.01 (repeated measures 2-way analysis of variance with Šídák’s post hoc test), n = 5. Abbreviations as in Figures 1 and 2.

Figure 5

Plasma levels of soluble GPVI Increase as CKD Advances, Whereas the Number of Platelet Granules as Well as CCL5 Content Are Reduced in CKD (A and B) Plasma levels of soluble GPVI as measured by enzyme-linked immunosorbant assay (ELISA). (A) Healthy control subjects n = 20, CKD3 n = 22, CKD4 n = 16, CKD5D n = 8, CKD5D ASA n = 11, CVD n = 4, and CVD ASA n = 10; (B) independent CARE FOR HOMe cohort with a total of 420 patients. (C) Plasma levels of 11-dehydro thromboxane B₂ as measured by ELISA, healthy control subjects n = 24, CKD3 n = 24, CKD4 n = 18, CKD5D n = 8, CKD5D ASA n = 10, CVD n = 5, and CVD ASA n = 10. (D) CCL5 levels in lysates of washed platelets of healthy control subjects (n = 14), CKD3 (n = 22), and CKD4 patients (n = 16) as determined by ELISA. (E) Average platelet granule count, number of granules per platelet, and representative images as analyzed by electron microscopy in platelets from 3 healthy control subjects and 5 CKD patients (CKD 4 n = 3, CKD5D ASA n = 2). Values are median with 25th-75th percentiles. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001 (Kruskal Wallis test with Dunn’s post hoc test). Abbreviations as in Figures 1 and 2.

Figure 6

Adenine-Fed Mice Present With Prolonged Bleeding Time, an Impaired Thrombotic Response In Vivo and Reduced In Vitro Thrombus Formation Under Flow ASA treatment restores the adenine-induced impairment of in vitro thrombus formation. (A) Schematic representation of the experimental set-up. (B) Bleeding was examined using the tail bleeding assay in which the time until the initial cessation of bleeding was monitored. n = 6-8 per group. (C) The in vivo thrombotic response was investigated upon a topical FeCl₃ application to arteries in the cremaster muscles. Time to full vessel occlusion was recorded. n = 8-9 per group. (B and C) When bleeding had not stopped (B) or the vessel was not occluded (C) at the end of the 20-minute analysis time, values were put to the maximum of 1,200 seconds. Values are median with 25th-75th percentiles. ∗P < 0.05; ∗∗∗P < 0.001 (Kruskal-Wallis test with Dunn’s post hoc test, outlier identification using the ROUT method). (D-F) PPACK anticoagulated mouse whole blood was perfused over a collagen type I surface at 1,000 s⁻¹ for 4 minutes. (D) Representative images of platelet adhesion and aggregation. Scale bar = 20 μm. Quantification of the surface area covered by adhered (E) and aggregated (F) platelets. Values are mean ± SD. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001 (1-way analysis of variance with Šídák’s post hoc test). n = 6-9 per group. Abbreviations as in Figures 1 and 2.