Germline genetic testing for patients with severe aplastic anemia (SAA) is recommended to guide treatment, including the use of immunosuppressive therapy and/or adjustment of hematopoietic cell transplantation (HCT) modalities. Hemophagocytic lymphohistiocytosis (HLH) is a life threatening hyperinflammatory condition often associated with cytopenias with autosomal recessive (AR) or X-linked recessive (XLR) inheritance. HLH is part of the SAA differential diagnosis and genetic testing may identify variants in HLH genes in patients with SAA. The impact of pathogenic/likely pathogenic (P/LP) variants in HLH genes on HCT outcomes in SAA is unclear. We aimed to determine the frequency of HLH gene variants in a large cohort of patients with acquired SAA and to evaluate their association(s) with HCT outcomes. The Transplant Outcomes in Aplastic Anemia project, a collaboration between the National Cancer Institute and the Center for International Blood and Marrow Transplant Research, consists of genomic and clinical data from 824 patients who underwent HCT for SAA between 1989 and 2015. We excluded 140 patients with inherited bone marrow failure syndromes and used exome sequencing data from the remaining 684 patients with acquired SAA to identify P/LP variants in 14 HLH-associated genes (11 AR, 3 XLR) curated using ACMG/AMP criteria. Deleterious variants of uncertain significance (del-VUS) were defined as those not meeting ACMG/AMP P/LP criteria but with damaging predictions in ≥3/5 meta-predictors (BayesDel, REVEL, CADD, MetaSVM and/or EIGEN). Kaplan-Meier estimator was used to calculate the probability of overall survival (OS) after HCT, and cumulative incidence calculator was used for other HCT outcomes accounting for relevant competing risks. There were 46 HLH variants in 49 patients (7.2%; N total=684). Seventeen variants in 19 patients (2.8%) were P/LP; 8 of these were loss of function variants. Among 19 patients with P/LP HLH variants, 16 (84%) had monoallelic variants in genes with AR inheritance, and three had variants in XLR genes. PRF1 was the most frequently affected gene (8/19 patients). We found no statistically significant differences in transplant-related factors between patients with and without P/LP HLH variants. The 5-year survival probabilities were 89% (95% CI=72-99), and 70% (95% CI=53-85%) in patients with P/LP and del-VUS HLH variants, respectively, as compared with 66% (95% CI=62-70) in those without variants (p-log-rank=0.16). The median time to neutrophil engraftment was 16 days for patients with P/LP HLH variants versus 18 days in those with del-VUS or without variants, combined (p-Gray's test=0.01). No statistically significant associations between P/LP HLH variants and the risk of acute or chronic graft-versus-host disease were noted. In this large cohort of acquired SAA, we found that 2.8% of patients harbor a P/LP variant in an HLH gene. No negative effect on post-HCT survival was noted with HLH gene variants. The small number of patients with P/LP HLH variants limit the study ability to provide conclusive evidence. Yet, our data suggest no need for special transplant considerations for patients with SAA carrying P/LP variants.
