SNP Karyotyping Can Detect Clonality of Stem Cell Compartment and Early Malignant Evolution in Aplastic Anemia

Aplastic anemia (AA) is characterized by pancytopenia due to contraction/destruction of stem cell compartment. Most investigators consider the presence of clonal cytogenetic abnormalities as incompatible with the AA diagnosis. Despite excellent response rates to immunosuppression (IS) in the majority of AA patients, clonal malignant evolution to myelodysplasia (MDS) can occur in 10–15% in 10 years. Among such cases monosomy 7 (mono7) is the most commonly reported cytogenetic abnormality.

Routine metaphase cytogenetics (MC) depends upon high numbers of dividing cells with inducible mitosis and is therefore often noninformative in AA. The inability to early detect AA patients at risk for clonal evolution constitutes a significant clinical problem. We hypothesized that high resolution SNP-array technology (SNPA) that allows for the analysis of interphase genomes will improve detection of chromosomal abnormalities in MDS evolving from AA. In addition to MC, we applied Affymetrix chips to study whole genomes of AA patients (N=100; 69 and 67 were investigated using 50K/250K and 6.0 arrays, respectively; 25 patients were studied at multiple time points prior and post IS). Data was analyzed using CNAG v3.0 and Genotyping Console v2.0 and unbalanced lesions were detected, including regions of genomic gain, loss and copy number neutral loss of heterozygozity.

Clonal malignant evolution was observed in 13 patients resulting in a conversion rate of 13%. We focused on longitudinal analysis of these patients. Abnormal endpoint MC was detected in a total of 69% of transformed patients (mono7 in 8/13, t(10;18)(q11.2;q21) in 1/13 and trisomy 12 in 1/13 patients, respectively). Remarkably in 5/13 (38%) of evolving AA patients, numerical aberrations were detected earlier by SNPA than MC. Mono7 (N=4) and trisomy 12 (N=1) were detected by SNPA in aspirates that showed normal or noninformative MC, while in subsequent analysis bone marrow exams concurred with the early diagnosis by SNPA. In 3 patients, MC and SNPA results were concordant and in 1 patient SNPA failed to early identify mono7,. Acquired uniparental disomy of various chromosomes was detected in a total of 4 patients and the analysis of nonmyeloid CD3+ cells revealed somatic nature of these lesions: 7q31.31–31.33(4.9mb), 17q11.2-qter(56mb), 6p12.1-pter(56mb) and 3q12.2-qter(97mb). Interestingly, 2 patients with UPD had normal concurrent MC and the clonal lesions were detected before clinical diagnosis of MDS was established. It is likely that depletion of stem cell compartment might favor detection of pseudoclonality manifested as non-pathogenic random chromosomal lesions. In our cohort, in 2 AA patients with normal MC, SNPA analysis prior to IS therapy revealed clonal UPD as confirmed by comparison of myeloid lineage with germ line configuration in CD3+ cells. These lesions disappeared post IS in both patients and nearly 2 years later a newly recruited pathogenic clone with mono7 was detected as a sole abnormality by SNPA and MC.

These patients are very illustrative as they may point towards genomic instability as well as depletion of available stem cell pool resulting in frequent recruitment of defective SC. Our study demonstrates that SNPA is a powerful tool to early identify AA patients harboring clonal defects consistent with the diagnosis of MDS, and thus it potentially might have strong clinical implications.