Patients with Waldenström macroglobulinemia have impaired platelet and coagulation function

Abstract

Clinical features in patients with the B-cell lymphoma, Waldenström macroglobulinemia (WM), include cytopenias, immunoglobulin M (IgM)-mediated hyperviscosity, fatigue, bleeding, and bruising. Therapeutics such as Bruton's tyrosine kinase inhibitors (BTKis) exacerbate bleeding risk. Abnormal hemostasis arising from platelet dysfunction, altered coagulation or vascular impairment have not yet been investigated in patients with WM. Therefore, this study aimed to evaluate hemostatic dysfunction in samples from these patients. Whole blood (WB) samples were collected from 14 patients with WM not receiving therapy, 5 patients receiving BTKis and 15 healthy donors (HDs). Platelet receptor levels and reticulation were measured by flow cytometry, plasma thrombin generation with or without platelets by fluorescence resonance energy transfer assay, WB clotting potential by rotational thromboelastometry, and plasma soluble glycoprotein VI (sGPVI) and serum thrombopoietin (TPO) by enzyme-linked immunosorbent assay. Donor platelet spreading, aggregation, and ability to accelerate thrombin generation in the presence of WM-derived IgM were assessed. WM platelet receptor levels, responses to physiological agonists, and plasma sGPVI were within normal ranges. WM platelets had reduced reticulation (P = .0012) whereas serum TPO levels were increased (P = .0040). WM plasma displayed slower thrombin generation (P = .0080) and WM platelets contributed less to endogenous thrombin potential (ETP; P = .0312). HD plasma or platelets incubated with IgM (50-60 mg/mL) displayed reduced spreading (P = .0002), aggregation (P < .0001), and ETP (P = .0081). Thus, alterations to thrombin potential and WB coagulation were detected in WM samples. WM IgM significantly impaired hemostasis in vitro. Platelet and coagulation properties are disturbed in patients with well-managed WM.

Graphical abstract

Figure 1.

WM platelets have altered thrombopoietic and age-related phenotypic markers. (A) Platelet count, (B) mean platelet volume (MPV), (C) serum thrombopoietin (TPO), (D) normalized thiazole orange (TO) staining of platelets in WB (divided by HD mean), (E) relative proportion of TO^bright WM platelets falling within the gate encompassing the top 10% of HD platelets (log₁₀ WM/HD %), (F) platelet surface receptor levels in PRP and (G) sGPVI levels in platelet-poor plasma (PPP) from HDs (n = 9-15, blue), patients with WM not on therapy (n = 9-13, red), or receiving BTKis (n = 0-5, green). For panels A-C,E,G, mean +/± SD. For panels D,F, normalized geomeans (divided by mean HD geomean) ± standard error of the mean (SEM). For panels A-C, P values were determined by Kruskal-Wallis 1-way analysis of variance (ANOVA) with Dunn's multiple comparisons test, performed on the raw data; for panel E, Mann-Whitney t test; or for panel E, 1 sample t test.

Figure 2.

WM platelets respond appropriately to standard agonists. (A-B) OgFg binding and (C-D) P-selectin exposure in WB and PRP samples, respectively, from HDs (n = 7-14, blue), patients with WM not on therapy (n = 8-13, red), or receiving BTKis (n = 0-5, green). Representative histograms from a HD showing (A) OgFg- or (C) P-selectin–positive events. Data (mean ± standard deviation [SD]) for (B) OgFg binding or (D) P-selectin exposure (%) in platelets treated with 0.5, 5 or 30 μg/mL crosslinked CRP-XL, 10 μM TRAP-6, or 5 μM ADP. P values were determined by Mann-Whitney t test with multiple comparisons.

Figure 3.

Patients with WM have delayed and reduced thrombin generation. Thrombin generation in (A-G) EDTA-anticoagulated PPP, (H) EDTA-anticoagulated PPP + PAR-1–activated HD WPs, or (I-M) trisoidium citrate-anticoagulated PPP + PAR1–activated HD or WM WPs in HDs (n = 6-25, blue), patients with WM not on therapy (n = 5-13, red) and receiving BTKis (n = 1-3, green). For panels A,H, representative thrombograms, mean of duplicates. For panels B-F,I-M, mean ± SD. For panel G, correlation between ETP and IgM levels of patients with WM. P values were determined by Kruskal-Wallis 1-way ANOVA with Dunn's multiple comparisons test for the panels B-F, or Wilcoxon matched-pairs signed rank test for the panels I-M. R² values were determined by simple linear regression.

Figure 4.

WM coagulation factors contribute robustly in a WB clotting assay. ROTEM measurements in WB samples from HDs (n = 31, blue), patients with WM not on therapy (n = 13, red) or receiving BTKis (n = 5, green). (A-C) INTEM is triggered by calcium, phospholipids, and ellagic acid. (D-F) EXTEM is triggered by calcium and tissue factor, whereas (G-I) FIBTEM additionally inhibits platelet cytoskeletal changes with cytochalasin D and reflects the coagulation factor contribution to thrombus formation. (J-K) EXTEM-FIBTEM (PLTEM) reflects the platelet contribution to thrombus formation. Mean ± SD. P values were determined by Kruskal-Wallis 1-way ANOVA with Dunn's multiple comparisons test. MCE, maximum clot elasticity (MCF × 100)/(100 − MCF); MCF, maximal clot firmness.

Figure 5.

High concentrations of WM-derived IgM impair platelet spreading, aggregation, and thrombin generation. HD WP spreading (n = 4-5) for 20-60 minutes on (A,C-F) fibrinogen or (B,G-J) collagen, then visualized for actin (red) and α-tubulin (green), in the presence of (C-F,G-J) vehicle (blue), BSA (equimolar to IgM, 4.11 mg/mL, pink), or IgM (60 mg/mL, red). (C,G) Number of adhered platelets; (D,H) platelet surface area; (E,I) circularity; or (F,J) aspect ratio, average of 5 field of views. (K) Representative aggregation trace, mean of duplicates. (L) WP aggregation responses to collagen (5 μg/mL), TRAP-6 (10 μM), and ADP (5 μM). (M) Representative thrombogram, mean of duplicates. Thrombin generation in HD samples (n = 8), with (N,R) lag time, (O,S) ETP, (P,T) peak thrombin generated, and (Q,U) thrombin generation velocity in PPP for panels N-Q or PPP + autologous WPs for panels R-U. For panel T, mean ± SD. P values were determined by 2-way ANOVA with multiple comparisons for panels C-F,G-J,L or Friedman test for panels N-U. AUC, area under the curve.