Initial Genomic Analysis of a Pure Erythroid Leukemia Developing in Association with Hydroyurea Treatment for Sickle Cell Anemia

Abstract 3254

The antimetabolite hydroxyurea (HU) is the only widely available therapy for reducing complications of sickle cell anemia. It is not widely prescribed due to concerns about adverse effects, including cancer. Insufficient data currently exist regarding its leukemogenic potential. HU associated leukemia in sickle cell anemia is exceptionally rare, with only 6 literature reports. Whether such cases are coincidental or related to HU remains unanswered. We identified a 33 year old male with pure erythroid leukemia (formerly M6 AML) and a unique combination of sickle cell anemia, HU treatment, and short treatment latency. Cytogenetics showed a complex karyotype with 5q rearrangement, 7q deletion, losses of chromosomes 15–22 and Y, suggesting a therapy related etiology. To date, no sickle cell anemia AML cases have undergone in-depth genomic analysis to determine if HU has an etiologic role in susceptibility to therapy related leukemogenesis. High density SNP genotyping was performed with 2.5 million markers to understand progressive chromosomal disease changes. Comparison of germline DNA prior to HU and tumor DNA identified 10 mosaic aneuploidy events present at AML diagnosis including losses at 5q, 7q and 17p and gains at 19p. Whole exome sequencing was also performed on the patient's germline, tumor and relapsed tumor DNA. We identified 33 putative somatic missense mutations including a validated somatic TP53 mutation, however somatic mutations were not present in commonly mutated AML genes like NPM1, RUNX1 or IDH1. Genome-wide cytogenetic band enrichment analysis of exome sequence data using the Gene Set Analysis Toolkit suggests the presence of a significant enrichment of mutations along chromosome 19p, and the same analysis of peripheral blood gene expression from an unrelated cohort of 10 sickle cell anemia subjects with brisk hematopoietic activity also suggested association of chromosome 19p with stress erythropoiesis. Together, the mosaic aneuploidy and exome sequence analysis converged on a region of 19p potentially important to the pathogenesis of pure erythroid leukemia and possibly stress erythropoiesis. This region extends at least 1.9 Mb and includes 58 genes. We are validating mutations in this region and exploring the role of 19p during stress erythropoiesis to further determine the significance of this finding. Specifically, several 19p mutations have been introduced into the BaF3 cell line to investigate their role in myeloid cell growth. Investigation of this rare leukemia case offers a unique opportunity to elucidate the pathogenesis of erythroleukemia and possibly to determine if HU therapy is associated with leukemogenesis.