Large gene panel sequencing in clinical diagnostics-results from 501 consecutive cases.

BACKGROUND: In addition to whole exomes, large gene panels of clinically associated genes are used as high-throughput sequencing tests in many clinical centers, but their clinical utility has been much less investigated. MATERIALS AND METHODS: Here we report the results of the 501 first unselected cases for whom TruSight One panel (Illumina Inc., San Diego, California) was sequenced as a clinical diagnostic test for a variety of indications in our department. The analysis was restricted to virtual subpanels based on referral forms, where doctors were asked to list candidate genes or select one from predefined larger panels. RESULTS: A probable or definite pathogenic finding was reported in 26.3% of cases. In 238 samples for whom 1 to 9 genes were requested for analysis, the diagnostic yield was significantly higher compared to other 263 cases for whom larger subpanels were requested (31.5% vs 21.7%, respectively, P = .016). Detected mutations included single nucleotide variants, small insertions and deletions, and larger copy number variants. Out of 157 reported mutations, 67 were previously undescribed. CONCLUSION: The clinical utility of large gene panel sequencing in the context of other genetic diagnostic tests is discussed in detail.

Factor V Leiden and prothrombin 20210G>A [corrected] mutation and paediatric ischaemic stroke: a case-control study and two meta-analyses.

AIM: To determine whether factor V Leiden (FVL) and prothrombin (PT) 20210G>A mutation are associated with paediatric ischaemic stroke. METHODS: The study consisted of two parts. Case-control study included neuroradiologically confirmed paediatric ischaemic stroke patients from two tertiary children's hospitals in Estonia. For control group, DNA was obtained from 400 anonymous screening test cards of newborns born consecutively in all delivery departments of Estonia in January 2005. Meta-analyses was performed to assess the association between paediatric sinovenous thrombosis and FVL and PT 20210G>A. RESULTS: A total of 75 children (45 boys, 30 girls) were included into the case-control study: 19 with childhood arterial ischaemic stroke, 49 with perinatal arterial ischaemic stroke and seven with cerebral venous thrombosis. Both FVL and PT 20210G>A occurred significantly more frequently among patients with sinovenous thrombosis compared with controls (OR = 12.9; 95% CI: 2.3-73.0 and OR = 11.9; 95% CI: 2.1-67.2, respectively). The difference was not significant between childhood/perinatal arterial ischaemic stroke and controls. Meta-analyses (including our study) revealed that both FVL and PT 20210G>A are associated with paediatric sinovenous thrombosis (OR = 3.1; 95% CI: 1.8-5.5 and OR = 3.1; 95% CI: 1.4-6.8, respectively). CONCLUSION: FVL and PT 20210G>A are associated with paediatric sinovenous thrombosis.

The number of CAG and GGN triplet repeats in the Androgen Receptor gene exert combinatorial effect on hormonal and sperm parameters in young men.

Androgen receptor (AR) is a transcription factor that is activated upon binding to testosterone (T) and is implicated in regulating the expression of reproduction-related genes. The human AR gene (Xq11-12) spans 186,588 bp and eight exons. N-terminal transactivation domain of the encoded AR protein harbours two polymorphic stretches of identical amino acids, a polyglutamine tract (encoded by 8-37 CAG-repeats) and a polyglycine tract (encoded by 10-30 GGN-repeats). We set forward to analyse independent and combinatory effects of the length of these repetitive tracts on male reproductive hormones, testicular and sperm parameters in a population-based cohort of Baltic young men (n = 974; aged 20.1 ± 2.1 years). We designed an assay to amplify and detect simultaneously the variants of both polymorphic repeats. The study revealed that elongated AR CAG tract was associated with lower FSH (linear regression: p = 0.0002, effect per repeat -0.056 IU/L). As a novel finding, the carriers of GGN-stretch with ≥24 repeats showed a trend for decreased sperm concentration (p = 0.027). Although neither of the variants exhibited an isolated effect on circulating T, their allelic combinations modulated serum T levels, as well as sperm concentration. The lowest T was measured for men carrying the AR gene with long CAG (n ≥ 25) and short GGN (n ≤ 21) repeat tracts (mean 18.8 vs. 25.5-28.6 nmol/L for the other AR variants, p = 0.017). The lowest sperm concentration was detected among individuals with both elongated repetitive stretches (CAG, n ≥ 25 and GGN, n ≥ 24; mean 49.0 vs. 68.4-72.1 mill/mL for the other variants; p = 0.00059). The innovative study design enabled to clearly demonstrate a combinatory impact of CAG and GGN repeat lengths at male reproductive parameters. As AR regulates transcription of over 900 genes in many tissues and organs, the combinatory effects of these common repeat-length variants on male physiology in the wider context and across lifetime are still to be assessed.

Hearing impairment in Estonia: an algorithm to investigate genetic causes in pediatric patients.

PURPOSE: The present study was initiated to establish the etiological causes of early onset hearing loss (HL) among Estonian children between 2000-2009. METHODS: The study group consisted of 233 probands who were first tested with an arrayed primer extension assay, which covers 199 mutations in 7 genes (GJB2, GJB6, GJB3, SLC26A4, SLC26A5 genes, and two mitochondrial genes - 12S rRNA, tRNASer(UCN)). From probands whose etiology of HL remained unknown, DNA analysis of congenital cytomegalovirus (CMV) infection and G-banded karyotype and/or chromosomal microarray analysis (CMA) were performed. RESULTS: In 110 (47%) cases, the etiology of HL was genetic and in 5 (2%) congenital CMV infection was diagnosed. We found mutations with clinical significance in GJB2 (100 children, 43%) and in 2 mitochondrial genes (2 patients, 1%). A single mutation in SLC26A4 gene was detected in 5 probands (2.2%) and was considered diagnostic. In 4 probands a heterozygous IVS2-2A>G change in the SLC26A5 gene was found. We did not find any instances of homozygosity for this splice variant in the probands. CMA identified in 4 probands chromosomal regions with the loss of one allele. In 2 of them we were able to conclude that the found abnormalities are definitely pathogenic (12q13.3-q14.2 and 17q22-23.2 microdeletion), but the pathogenity of 2 other findings (3p26.2 and 1p33 microdeletion) remained unknown. CONCLUSION: This practical diagnostic algorithm confirmed the etiology of early onset HL for 115 Estonian patients (49%). This algorithm may be generalized to other populations for clinical application.

Prevalence of Long-Chain 3-Hydroxyacyl-CoA Dehydrogenase Deficiency in Estonia.

The aim of our study was to evaluate the prevalence of long chain 3-hydroxyacyl-CoA dehydrogenase deficiency (LCHADD) in the general Estonian population and among patients with symptoms suggestive of fatty acid oxidation (FAO) defects. We collected DNA from a cohort of 1,040 anonymous newborn blood spot samples. We screened these samples for the presence of the common c.1528G>C mutation in the HADHA gene. Based on the clinical suspicion of FAO defects, we screened suspected individuals since 2004 for the common c.1528G>C mutation in the HADHA gene and since 2008 in addition by tandem mass spectrometric analysis of plasma acylcarnitines. Our results showed that the carrier frequency of the c.1528G>C mutation in the Estonian population is high - 1:173. During the screening of symptomatic patients, we identified five LCHADD patients in four families. Three patients were retrospectively identified by molecular screening of the HADHA gene. One patient was homozygous for the c.1528G>C mutation in the HADHA gene, and two siblings were compound heterozygotes with HADHA genotype c.[1528G>C]+[1690-2A>G]. Among patients tested using acylcarnitine profiling, we identified two cases with an abnormal acylcarnitine profile typical to LCHADD. Molecular analysis showed homozygosity for c.1528G>C mutation. Based on a carrier frequency of 1:173 (95% Confidence Interval 1:76-1:454) and taking into account that the c.1528G>C mutation makes up 87.5% of disease alleles in Estonian LCHADD patients, the estimated prevalence of LCHADD in Estonia would be 1: 91,700.