HLA-C KIR-Ligands Determine the Impact of Anti-Thymocyte Globulin (ATG) on Graft versus Host and Graft versus Leukemia Effects Following Hematopoietic Stem Cell Transplantation

摘要

Rabbit anti-thymocyte globulins (ATGs) are widely used for the prevention of acute and chronic graft versus host disease (aGVHD, cGVHD) following allogeneic hematopoietic stem cell transplantation (HSCT). However, most prospective and retrospective studies did not reveal an overall survival (OS) benefit associated with ATG. Homozygosity for human leukocyte antigen (HLA)-C group 1 killer-cell immunoglobulin-like receptor ligands (KIR-L), i.e. C1/1 KIR-L status, was recently shown to be a risk factor for severe aGVHD. Congruously, we have previously reported favorable outcomes in C1/1 recipients after ATG-based transplants in a monocentric analysis. Here, within an extended cohort, we test the hypothesis that incorporation of ATG for GVHD prophylaxis may improve survival particularly in HSCT recipients with at least one C1 KIR-ligand. Retrospectively, 775 consecutive allogeneic (excluding haploidentical) HSCTs were analyzed, including peripheral blood and bone marrow grafts for adults with hematological diseases at two Austrian HSCT centers. ATG-Fresenius/Grafalon, Thymoglobuline, and alemtuzumab were applied in 256, 87, and 7 transplants, respectively (subsequently summarized as “ATG”), while 425 HSCT were performed without ATG. Median follow-up of surviving patients is 48 months. Adjusted for age, disease-risk, HLA-match, donor and graft type, sex match, cytomegalovirus serostatus, conditioning intensity, and type of post-grafting GVHD prophylaxis, Cox regression analysis of the entire cohort ( n = 775) revealed a significant association of ATG with decreased non-relapse mortality (NRM) (risk ratio (RR), 0.57; p = 0.001), and overall mortality (RR, 0.71; p = 0.014). Upon stratification for HLA-C KIR-L, the greatest benefit for ATG emerged in C1/1 recipients ( n = 291), by reduction of non-relapse (RR, 0.34; p = 0.0002) and overall mortality (RR, 0.50; p = 0.003). Less pronounced, ATG decreased NRM (RR, 0.60; p = 0.036) in HLA-C group 1/2 recipients ( n = 364), without significantly influencing overall mortality (RR, 0.70; p = 0.065). After exclusion of higher-dose ATG-based transplants, serotherapy significantly improved both NRM (RR, 0.54; p = 0.019; n = 322) and overall mortality (RR, 0.60; p = 0.018) in C1/2 recipients as well. In both, C1/1 (RR, 1.70; p = 0.10) and particularly in C1/2 recipients (RR, 0.94; p = 0.81), there was no statistically significant impact of ATG on relapse incidence. By contrast, in C2/2 recipients ( n = 121), ATG neither reduced NRM (RR, 1.10; p = 0.82) nor overall mortality (RR, 1.50; p = 0.17), but increased the risk for relapse (RR, 4.38; p = 0.02). These retrospective findings suggest ATG may provide a survival benefit in recipients with at least one C1 group KIR-L, by reducing NRM without significantly increasing the relapse risk.