Systemic complement activation influences outcomes after allogeneic hematopoietic cell transplantation: A prospective...

Complement activation has shown a role in murine models of graft-versus-host disease (GVHD) and in endothelial complications after allogeneic hematopoietic cell transplantation (allo-HSCT). However, its impact on post-transplant outcomes has not been so far fully elucidated. Here, we conducted a prospective multicentric trial (NCT01520623) performing serial measurements of complement proteins, regulators, and CH50 activity for 12 weeks after allo-HSCT in 85 patients receiving a myeloablative conditioning (MAC) regimen for various hematological malignancies. Twenty-six out of 85 patients showed an "activated" complement profile through the classical/lectin pathway, defined as a post-transplant decline of C3/C4 and CH50 activity. Time-dependent Cox regression models demonstrated that complement activation within the first weeks after allo-HSCT was associated with increased non-relapse mortality (hazard ratio [HR]: 3.69, 95% confident interval [CI]: 1.55-8.78, p = .003) and poorer overall survival (HR: 2.72, 95% CI: 1.37-5.39, p = .004) due to increased incidence of grade II-IV acute GVHD and in particular gastrointestinal (GI) GVHD (HR: 36.8, 95% CI: 12.4-109.1, p < .001), higher incidences of thrombotic microangiopathy (HR: 8.58, 95% CI: 2.16-34.08, p = .0022), capillary leak syndrome (HR: 7.36, 95% CI: 2.51-21.66, p = .00028), post-engraftment bacterial infections (HR: 2.37, 95% CI: 1.22-4.63, p = .0108), and EBV reactivation (HR: 3.33, 95% CI: 1.31-8.45, p = .0112). Through specific immune staining, we showed the correlation of deposition of C1q, C3d, C4d, and of C5b9 components on endothelial cells in GI GVHD lesions with the histological grade of GVHD. Altogether these findings define the epidemiology and the clinical impact of complement classical/lectin pathway activation after MAC regimens and provide a rational for the use of complement inhibitory therapeutics in a post-allo-HSCT setting.