Implementation of the diagnostic workflow of hereditary anemias by using targeted-ngs panel analysis.

Abstract

Background: Mutations in more than 70 genes cause hereditary anemias (HA), a highly heterogeneous group of rare/low frequency disorders in which we included: hyporegenerative anemias, as congenital dyserythropoietic anemias (CDA) and Diamond-Blackfan anemia; erythrocyte membrane defects, as hereditary spherocytosis and stomatocytosis; hemolytic anemias due to enzymatic defects, as pyruvate kinase (PK) deficiency. The classification and the distinction among different types of HA is often difficult. The variety of unspecific and overlapping syndromic and non-syndromic phenotypes somewhat hampers a clear clinical diagnosis and prevents straightforward genetic testing. The diagnosis of these conditions may require several lines of lines of investigation. Due to the failure of the current diagnostic workflow to find a definitive and correct diagnosis of HA, next-generation sequencing (NGS) is making its way on this field. The major current application of NGS in diagnostics is through targeted (t)-NGS, in which a selected fraction of genes is analyzed (custom gene panels). Aim(s):We propose a new diagnostic workflow for HA based on t-NGS. Method(s): We have developed two consecutive versions of a t-NGS panel, including 34 and 71 genes, respectively. Seventy-four probands from 62 unrelated families were investigated. The probe design was performed by web-based tool SureDesign (Agilent Technologies, USA). Sample preparation was performed following the instruction's manufacturer for HaloPlex Target Enrichment kit for Illumina Sequencing (Agilent Technologies). Highthroughput sequencing was performed by Illumina NextSeq 500. Agilent SureCall software (v 3.0.3.1, Agilent Technologies) was used for bioinformatic and computational analyses. The pathogenicity of each variant was evaluated by gathering evidence from various sources, according to the guidelines of ACMG. Result(s):We obtained an overall diagnostic yield of 64.9%. Despite 54.2% of cases showed conclusive diagnosis fitting well to the clinical suspicion, the multi-gene analysis modified the original clinical diagnosis in 45.8% of patients (non-matched phenotype-genotype). Of note, 81.8% of nonmatched patients were clinically suspected to suffer from CDA. Particularly, 45.5% of the probands originally classified as CDA exhibited a conclusive diagnosis of chronic anemia due to enzymatic defects, mainly due to mutations in PKLR gene. Interestingly, we also identified a syndromic CDA patient with mild anemia and epilepsy, showing a homozygous mutation in CAD gene, recently associated to early infantile epileptic encephalopathy- 50 and CDA-like anemia. Multi-target diagnosis also allowed us identifying polygenic conditions, in which the phenotypic variability could be explained by the co-inheritance of multiple disease mutations. This was the case of a PIEZO1 patient that also exhibited the E109K-SEC23B genotype, causative of CDAII. The co-inheritance of PIEZO1 and SEC23B mutations accounts for marked iron overload in this patient. Summary and Conclusion(s): We herein described the diagnostic workflow that we established for molecular diagnosis of HA, based on the development of two consecutive versions of a t-NGS panel, including 34 and 71 genes, respectively. We showed the results obtained by the analysis of 74 probands. This is the largest cohort of patients and the most comprehensive gene set for this subset of anemias described so far. We also demonstrated that the multi-gene approach is valuable not only for achieving a correct and definitive diagnosis, but also for guiding treatment.