Comprehensive Molecular Profiling of Idiopathic Erythrocytosis By Genetic Ancestry

Introduction: Idiopathic erythrocytosis (IE) is characterized by a persistently elevated hemoglobin, equivocal erythropoietin (EPO) levels, absence of janus kinase 2 (JAK2) mutations suggestive of polycythemia vera (PV) and no secondary cause. One study used a targeted 21 gene next-generation sequencing panel and identified novel variants in known erythrocytosis-related genes as well as novel genes associated with the oxygen-sensing pathway. However, expanded sequencing of blood and matched tissue samples in a large ethnically diverse group of IE patients has not been performed.

Methods: All patients signed informed consent to participate in an observational study approved by the Institutional Review Board; they provide blood and buccal mucosa samples at study entry and at 24-month follow-up. Patients were enrolled if JAK2 testing and a complete work up for secondary causes was negative. They were required to have hemoglobin levels greater than 16 g/dL on two occasions or greater than 15 g/dL if undergoing phlebotomy. Our initial sequencing of 20 IE patients was performed utilizing high resolution whole-exome sequencing of circulating blood samples (disease) at a mean coverage of 390x and matched normal (buccal) samples at mean coverage of 300x. To stratify samples by genetic ancestry, we performed a population stratification principle component analysis (PCA) and STRUCTURE using Ancestry Informative Markers derived from 1K Genome Phase1_v3 Exome database. The primary in-silico analysis was performed on the baseline samples from treatment-naïve patients. The whole-exome data was generated in accordance to GATK's best practices with same filters applied as described by Exome Aggregation Consortium. The additional downstream in-silico paired analysis was performed using MutSig2.0 (Mutation Significance) algorithm to determine significant mutations and GISTIC (The Genomic Identification of Significant Targets in Cancer) to identify the significant copy number events, IPA (Ingenuity Pathway Analysis) to determine pathways along with other computational .

Results: Median age at baseline was 52 years (range 35-71). Six patients (30%) were female and 14 (70%) were male. Median values and ranges for laboratory parameters at baseline were as follows: WBC 6.6 x 10⁹/L (5-9.7), Hgb 17 g/dL (15.5-19.8), Plt 218 x 10⁹/L (86-374), and EPO level 9.8 IU/L (2-14.3). Three patients had a personal history of malignancy, including 2 with lymphoma. Two patients had a family history of myeloid malignancy (chronic myeloid leukemia and PV). Our ancestry analysis of initial 20 patients with IE identified 6 patients with high European percent ancestry (EUR), 1 patient with high Asian percent ancestry (EAS) and 13 patients with high percent Ad Mixed ancestry (AMR). In our cohort, 60% (12/20) of patients had been also diagnosed with a liver disorder (11 with fatty liver, 1 with cirrhosis) that was not significantly different across populations. We identified, on average, 42 non-silent somatic mutations (not present in the buccal samples) in whole blood across our cohort with no statistical difference (p=0.671) in mutation burden between ancestry groups or between patients with and without liver disease. Age, gender, and ethnicity were not associated with mutation burden. Utilizing MutSig algorithm, we identified a novel candidate gene, CHAF1A, with high mutation prevalence of 30% in patients with IE. Further analysis of mutation landscape identified somatic nonsilent mutations in 25 known oncogenes which were present in at least 10% of patients. Our mutation signatures in IE identified a significant association with failure of double stranded DNA repair. Only one patient had a mutation in TET2. Further analysis of copy number indicated copy number loss in genes such as SETD3 and GSH associated with chromatin assembly which may suggest alterations in chromatin assembly and changes in the epigenome. Our analysis also identified a high number of 9p and 13q gains in patients with IE.

Conclusion: In this study, we utilized high-resolution next generation sequencing in association with comprehensive clinical annotation to determine potential molecular drivers of IE in a multi-ethnic population. We identified somatic mutations in a subset of patients which may represent clonal hematopoiesis. Long term follow up of outcomes in this cohort may clarify the significance of these mutations in the pathogenesis of IE.