Introduction Acquired aplastic anemia (aAA) occurs in both children and adults, though pathophysiology and clinical outcomes differ significantly by age. While clinical testing for paroxysmal nocturnal hemoglobinuria (PNH) clones is routine, targeted next generation sequencing (NGS) for myelodysplastic syndrome (MDS) associated variants has only recently been implemented. Furthermore, single nucleotide polymorphism array (SNP-A) assessing for copy neutral loss of heterozygosity in the human leukocyte antigen (HLA) region on chromosome 6p (6p CN-LOH) and HLA NGS assessing for allele specific loss of function (LoF) mutations, which together are seen in up to 25% of aAA cases, are only now being considered for clinical use. Recent studies have reported on the frequency and overlap of these 3 types of clonal evolution (PNH, MDS-associated, HLA-targeting) in adult aAA cohorts, but none focus specifically on pediatric patients, whose MDS-associated clonal evolution patterns are distinct. Delineating patterns of clonal hematopoiesis in pediatric aAA to determine whether evolution driven by immune escape (PNH and HLA loss) affects likelihood of MDS development is critical for personalizing therapeutic options. Thus, we aimed to describe the frequency, clinical significance, and overlap of clonal evolution including PNH, HLA loss, and MDS-associated gene mutations in a pediatric cohort of aAA patients. Methods This study identified 85 pediatric aAA patients treated at Children's Hospital of Philadelphia (CHOP) and enrolled on the IRB-approved CHOP/Penn BMF Registry and Repository in whom clinical testing for all patterns of clonal hematopoiesis was available or could be performed with biorepository samples. PNH flow cytometry was performed clinically in all cases; subjects with PNH > 0.05% in the granulocyte fraction were considered as having a PNH clone. Peripheral blood and bone marrow samples were used for SNP-A, HLA-NGS, and an institutional somatic mutation NGS panel for hematological malignancies encompassing 118 genes. Descriptive and contingency statistical analyses along with Fisher's Exact Test were performed using SAS (Cary, NC). Results Demographic composition of the 85 pediatric subjects with aAA included a median age of 10.7 years (1.7-20.8), male/female ratio of 59%/41%, and a representative race/ethnicity distribution (Table 1). In total, 66% displayed at least one clonal event. As expected, PNH clones were most common, present in 45% (38/85) of the cohort. An additional 16 (19%) exhibited acquired HLA loss by 6p CN-LOH (n=6), allele specific LoF mutations (n=6), or both (n=4). Fifteen (18%) had clonal hematopoiesis associated with MDS, of whom 3 had cytogenetic lesions that progressed to clinical MDS (3.5%), and 12 developed no morphologic features of MDS but demonstrated a total of 14 mutations in MDS-associated genes (BCOR/BCORL1 n=6, KMT2C/KMT2D n=3, RUNX1 n=1, ASXL1 n=1, SRSF2 n=1, PRPF8 n=1, CALR n=1). Critically, pediatric subjects with aAA who developed HLA loss did not develop MDS-associated mutations, with no patient demonstrating both subtypes of clonal events (Figure 1). In contrast, PNH and MDS-associated clonal events overlapped in 11% of the overall cohort. In total, 60% of patients with MDS-associated clonal hematopoiesis also possessed PNH clones, and 23% of patients with detectable PNH clones demonstrated at least one MDS-associated clonal event. Only 3.53% (n=3) of the cohort demonstrated both PNH and HLA loss, and PNH clone size was < 1% in all 3 patients. Furthermore, patients with HLA loss were significantly less likely to have PNH clones than those without HLA somatic alterations (19% vs 51%, p = 0.026). Correlations between patterns of clonal hematopoiesis and clinical outcomes will be presented. Concalusions Pediatric aAA patients treated with IST are at risk for developing MDS and hemolytic PNH. Presence and evolution of clonal hematopoiesis influenced decisions to pursue IST versus alternative donor transplant. Importantly, our study demonstrates that occurrence of clonal hematopoiesis by HLA loss does not increase and may even decrease the risk of developing MDS-associated or PNH clones. This finding strongly suggests that treatment decisions for aAA patients whose only clonal event is HLA loss need not be considered distinctly from patients with no clonal changes. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal