Complement Activation Is Associated with Endothelial Damage in Hematopoietic Stem Cell Transplant Associated-Thrombotic Microangiopathy

Abstract

Introduction: Hematopoietic stem cell transplant-associated thrombotic microangiopathy (HSCT-TMA) is a life-threatening complication of HSCT. Identification of at-risk patients is challenging, leading to delays in diagnosis and poor outcomes. Degradation of the glycocalyx and dysregulated complement activation are integral to the pathophysiology. The glycocalyx components thrombomodulin (TM) and syndecan-1 (SYND1) are shed during cell damage, and detection in patient plasma may serve as biomarkers of endothelial cell (EC) dysfunction. Therapeutic intervention to preserve glycocalyx integrity may protect the endothelium and also improve the clinical outcome during HSCT-TMA. Additionally, plasma Ba, an alternative complement pathway (AP) activation product, has been identified as a prognostic biomarker of HSCT-TMA. Our objective was to explore the link between excessive AP activation and endothelial damage in the context of HSCT-TMA and to potentially identify biomarkers to aid in its diagnosis. Methods: Biomarker levels of soluble TM, SYND1, and Ba were measured in plasma of pediatric HSCT patients with and without (controls) TMA obtained from Texas Children's Hospital. Sample acquisition was approved by Baylor College of Medicine Institutional Review Board. Biomarkers were also measured in a pre-clinical in vivo model of inflammation-mediated complement activation similar to HSCT-TMA, in which C57BL/6J mice were injected with lipopolysaccharide (LPS) to induce complement activation, followed by treatment with AP (anti-fB IgG) and terminal complement pathway (anti-C5 IgG) inhibitors or an IgG isotype control. Finally, as the immunosuppressant cyclosporin A (CsA) is a known trigger of HSCT-TMA, the role of complement in CsA-mediated damage in human ECs (HMEC-1 cells and HUVECs) was assessed in vitro. ECs were incubated with CsA followed by normal human serum (NHS) in the presence or absence of the C5 complement inhibitor eculizumab. TM was then measured by fluorescent staining on the cell surface and quantified using mean fluorescence intensity (MFI). Results: HSCT-TMA patients (n=11), compared to HSCT control patients (n=7), had significantly elevated levels of biomarkers of glycocalyx damage (TM: 18.3 ± 10.9 ng/mL and 6.6 ± 2.9 ng/mL, respectively; P<0.01; SYND1: 175.3 ± 177.0 ng/mL and 35.8 ± 18.9, respectively; P<0.05). A strong positive correlation between TM and SYND1 was observed (Pearson r=0.89). Plasma Ba was also significantly elevated in patients with TMA compared to those without (1790.0 ± 1262.0 ng/mL and 652.1 ± 233.5 ng/mL, respectively; P<0.05), and correlated positively with both TM and SYND1 levels (Pearson r=0.72 and 0.50, respectively). We have previously shown in an in vivo model of inflammation-mediated complement activation that LPS-injected mice, compared to saline-injected control mice, had significantly increased plasma Ba levels. Here we demonstrated LPS-injected mice (which also received IgG isotype control) had higher circulating levels of TM than untreated mice (35.1 ± 5.8 ng/mL [n=6] and 10.9 ± 1.6 ng/mL [n=6], respectively; P<0.0001). Inhibition of alternative (anti-fB IgG) and terminal (anti-C5 IgG) pathways of complement significantly attenuated TM levels (compared to isotype control [35.1 ± 5.8 ng/mL]; anti-FB: 25.2 ± 5.3 ng/mL, P<0.01; anti-C5: 26.2 ± 4.9 ng/mL, P<0.05, Fig. 1). TM was positively correlated with Ba (Pearson r=0.68). In vitro treatment of HMEC-1 cells with CsA led to deposition of iC3b and C5b-9, and C5 inhibition with eculizumab significantly reduced C5b-9 deposition (P< 0.0001) but had no effect on iC3b deposition. Treatment of HUVECs with CsA also reduced surface expression of TM (P<0.0001), and TM loss was partially restored by eculizumab (P<0.05, Fig 2). Conclusion: Plasma biomarkers of endothelial damage were elevated in patients with HSCT-TMA and in mice with inflammatory-mediated complement activation. Furthermore, AP activation positively correlated with biomarkers of endothelial damage. Blockade of complement activation significantly reduced markers of endothelial damage in both in vivo and in vitro models. These results provide further evidence linking complement activation and endothelial damage in HSCT-TMA and suggest that the use of these biomarkers may aid in the diagnosis of HSCT-TMA. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal