Identification of Recurrent Submicroscopic Deletions of chromosome1q31.3 As Possible Molecular Prognostic Markers in Acute Promyelocytic Leukemia (APL)

The characteristic hallmark of acute promyelocytic leukemia (APL) is a balanced reciprocal chromosomal translocation between chromosomes 15 and 17 leading to a fusion product consisting of the promyelocytic leukemia gene (PML) and the retinoic acid receptor alpha (RARA). The PML-RARA fusion product is necessary but not sufficient for the generation of leukemia and it is hypothesized that additional genomic lesions play a role in the pathogenesis of APL. Therefore, we have performed a high density SNP array analysis on 101 APL patient samples to identify new copy number alterations (CNAs) which may be relevant for the biology and prognosis of APL. Patients and Methods: Molecular and clinical outcome analyses were carried out retrospectively on patients diagnosed with APL, whose samples were referred to the molecular laboratory of the Department of Hematology and Oncology of the Medical Faculty Mannheim, University of Heidelberg, Germany between 1997 and 2010. 500 ng of genomic DNA from leukemic blasts per sample were processed according to the Genome Wide Human SNP 6.0 Array protocol (Affymetrix, Santa Clara, CA). The CNAG 3.3 software was used to perform allele-specific copy number analysis with anonymous references. CNAs of special interest were validated to be acquired in leukemia cells by performing allele-specific copy number analysis in matched pair SNP 6.0 array analysis, quantitative real time PCR and direct sequencing of genomic DNA from initial diagnosis and molecular remission samples. Results: We identified 279 acquired CNAs consisting of 185 heterozygous deletions, 87 amplifications and 7 regions of copy number neutral loss of heterozygosity (CNLOH). Besides common chromosomal aberrations such as trisomy 8 or duplications of the long arm of chromosome 8, deletions of 7q or isochromosome ider(17)(q10)t(15;17), numerous novel recurrent micro-deletions were discovered. The most common was a somatically acquired ∼100 kilobase deletion of chromosome 1q31.3 in 13 of 101 (13%) patients. These deletions encompassed the coding regions for the microRNAs mir181a1/b1. In univariate analyses of overall survival (OS) and relapse free survival (RFS) using Logrank tests, we found that patients carrying 2 or more CNAs as compared to 0 or 1 CNA as detected by SNP array had a significantly increased risk of death (p=0.016) and relapse (p=0.019). Patients carrying the recurrent deletion of chromosome 1q31.3 as compared to patients not carrying this deletion had a significantly increased risk of relapse (p=0.005), a markedly higher number of CNAs (median 8 vs. 2, p < 0.0001) and significantly higher white blood cell counts (WBC) at initial diagnosis (median 2550/μl vs. 16900/μl, p=0.009). We performed a multivariable analysis using Cox proportional hazards models to evaluate power of CNAs detected by SNP-array and deletions of chr. 1q31.3 as possible independent prognostic markers in APL as compared to age, WBC, platelet count (PC) and FLT3 mutational status. For the full model of OS only age and the number of CNAs detected by SNP-A met a 0.10 level of entry into the model and confirmed that the group of patients with 2 or more CNAs represent a subgroup with inferior outcome (“2 or more lesions”: hazard ratio = 5.942, p = 0.0015, age: hazard ratio = 1.08, p < 0.0001). Of note, for the analysis of RFS the presence of a chr. 1q31.3 deletion was the only effect, which met the entry level into the model of RFS and therefore was identified as a new strong predictor for an increased risk of relapse (“presence of del1q31.3”: hazard ratio = 28.9, p = 0.0031). Conclusion: The profiles of submicroscopic CNAs in APL patients are heterogeneic and may serve as strong independent prognostic markers for disease risk definition. Recurrent submicroscopic deletions of chr. 1q31.3 in leukemia cells of APL patients were associated with an increased number and characteristic pattern of further newly identified CNAs, unfavorable laboratory parameters and a higher risk of relapse. The number of CNAs was shown to be predictive for early death rate and overall survival. The further pursuit of these new potential molecular markers is highly warranted as they could refine the current risk stratification of APL by identifying new subgroups of patients, who could possibly gain from adapted treatment strategies.

No relevant conflicts of interest to declare.