Recurrent 6pLOH Is the Most Common Somatic Lesion in Refractory Cytopenia of Childhood and Occurs Very Infrequently in Severe Aplastic Anemia

Abstract 644 Refractory cytopenia of the childhood (RCC) and severe aplastic anemia (SAA) are the most common causes of acquired hypoplastic bone marrow failure (BMF) in children. Although predisposing genetic factors and inciting immunological events had been implicated, little is known about the molecular origins of these conditions. Whole-genome scanning using single nucleotide polymorphism arrays (SNP-A) can complement standard cytogenetics in terms of the detection of submicrocopic aberrations and regions with copy number neutral loss of heterozygosity (CN-LOH). Here we present the results of an analysis of 181 children with bone marrow failure: MDS, n=106; SAA, n=41; and for hematological control, 34 patients with Diamond Blackfan anemia (DBA). To enrich for the myeloid lineage, we used granulocytes from bone marrow as source for DNA. We employed Affymetrix 6.0 arrays and used CNAG v3.3 and Genotyping Console v4.0 platforms for data analysis. To identify somatic aberrations that might have been missed by standard metaphase cytogenetics (MC), we initially looked at cases with abnormal MC, presenting as RCC or advanced MDS (n=25, median age 12.4 years). While SNP-A generally confirmed the cytogenetic lesions identified by MC, no novel recurrent somatic aberrations with pathogenic character were discovered. In the next step we focused on the analysis of RCC patients with normal karyotype (N=81, median age 10.2 years). In one case a small monosomy 7 clone was identified using SNP-A, that has been missed by standard MC. Most strikingly, 11 patients (14%) carried clones with a terminal CN-LOH of the short arm of chromosome 6 (6pLOH), with a length of 30–42Mb and various clonal size. The somatic myeloid origin was confirmed by the analysis of CD4+/CD8+ sorted T-cells in several cases. The 6pLOH lesion encompasses the HLA gene cluster leading to the loss of one HLA haplotype, has previously been reported in adults with SAA diagnosed in USA and Japan (Afable/Wlodarski, and Katagiri 2011), and was generally associated with a good prognosis. Regarding treatment modalities in RCC patients, the 6pLOH clone was overrepresented in the watch and wait (W&W) cohort: 10/55 (18%) of W&W cases carried this clone, as compared to 1/26 (4%) patient who received therapy within 6 months from diagnosis. Out of the 10 W&W patients with 6pLOH clone, only two required later therapeutic interventions (stem cell transplantation) due to progressive cytopenia at 10 and 11 months after diagnosis. The remaining 8/10 patients are still under W&W strategy with a median FUP of 6.6 (1.6.-12.2) years. In summary, all 11 RCC 6pLOH patients are alive with a median FUP of 6.6 years without disease progression. Interestingly, the 6pLOH clone survives over a long period of time, as confirmed in 2 patients in different bone marrow samples obtained 7 years apart. To answer the question if the persistence of the 6pLOH clones has an effect on telomere maintenance in the hematopoietic cell compartment, we measured telomere length using a quantitative PCR-based approach. All 11 RCC cases with 6pLOH studied had normal telomere length as compared to age-matched controls. Since the immune-mediated attack is the main operating mechanism of BMF in SAA, we next asked whether the 6pLOH clone can also arise in the bone marrow of children with SAA. We SNP-A genotyped a cohort of 41 SAA patients (median age 10.4 years) including 5 hepatitis-associated SAA (HSAA) cases. While no genomic copy number alterations were found, the somatic 6pLOH clone was discovered in only one HSAA patient, in whom the SAA developed 6 months after onset of hepatitis. When compared to the SAA patients, the 6pLOH clone is significantly more frequent in the RCC W&W cohort (P In summary, 6pLOH is the most common somatic lesion found in the myeloid compartment in RCC patients and correlates with a very good prognosis. The absence of 6pLOH in children with SAA with no other associated pathologies supports the concept that RCC and SAA are two distinct entities in children. Disclosures: No relevant conflicts of interest to declare.