Accumulating evidence suggests that, if they go undetected, somatic low-level mutations in heterogeneous tumors or pre-cancerous syndromes can have profound clinical consequences. Furthermore, they may comprise important biomarkers and 'missed opportunities' for optimized therapy that could be applied had the mutation been known. We shall examine this hypothesis for the specific case of myelo-dysplastic syndromes (MDS). MDS are a collection of pre-cancerous, clonal bone marrow disorders with an increased risk of progression to acute myeloid leukemia (AML). Collaborators Ebert and Bejar demonstrated that mutations in TP53, RUNX1, ASLX1, EZH2, ETV6, are associated with decreased overall survival and are independent prognostic factors of outcome in multivariate analysis. Overall, mutations in MDS patients are of growing significance as biomarkers for 'personalization' of therapy beyond the established International Prognostic Scoring System (IPSS) which is based on clinical features and cytogenetics. While MDS is clearly a genetically heterogeneous disease, it is still not clear whether rare sub-clones influence clinical phenotype. This study aims first, to determine the prevalence of specific mutations that are below detection by existing technologies, and second to determine whether any identified mutations alter clinical phenotype. We shall employ COLD-PCR, a method developed by our group for enriching and detecting low-level DNA mutations, in conjunction with amplicon-based next-generation-sequencing. COLD-PCR increases the sensitivity of Illumina-based amplicon sequencing from the current ~2-5% down to 0.04% abundance, i.e. 'deep-sequencing' becomes 'ultra-deep-sequencing' using COLD-PCR. DNA from a group of 287 MDS patients will be screened via COLD-PCR-Illumina for mutations in the prognostic/potentially prognostic genes. Data from two groups of patients (poor outcome vs. favorable outcome, similar IPSS score) will be analyzed (a) accounting for both, low-level (0.04- 5% abundance) and high level (>5% abundance) mutations; and (b) accounting only for high level mutations, as practice has been until now. Results will be compared for their correlation with outcome/survival. The revised application contains additional data that fully validate our hypothesis for the first gene examined, NRAS. The ability to identify prognostic low-level mutations in MDS patients will enable better prediction of outcome and the fine-tuning of treatment for these patients. The approach addresses a problem common to all heterogeneous tumors (e.g. lung, pancreatic CA). Therefore relevance to Public Health is high.
Accumulating evidence suggests that, if they go undetected, somatic low-level mutations in heterogeneous tumors or pre-cancerous syndromes can have profound clinical consequences; and furthermore, they may comprise important biomarkers and 'missed opportunities' for optimized therapy that could be applied had the mutation been known. We shall examine this hypothesis for the specific case of myelo-dysplastic syndromes (MDS). MDS are a collection of pre-cancerous, clonal bone marrow disorders characterized by inefficient hematopoiesis, low blood counts, and an increased risk of progression to acute myeloid leukemia (AML). Somatic mutations in MDS patients are of growing significance as biomarkers that can refine personalization of therapy beyond the capability of established cytogenetic approaches. Using a newly developed technology (COLD-PCR) in combination with next generation sequencing we plan to uncover the existence and potential influence of occult mutations that can be used to fine-tune the treatment of MDS patients. The approach is generally applicable to many cancers.
