Rare monogenic severe prenatal or paediatric onset disorders are frequently both phenotypically and genetically heterogeneous, making these disorders very difficult to diagnose. 

Exome sequencing can be used as a first line test for patients with a clinical presentation indicative of a monogenic disorder where it is likely the most cost-effective strategy or for patients where standard genetics tests have not identified a diagnosis.

Our rapid trio “whole” exome sequencing service aims to provide a diagnosis for children with severe paediatric disorders who are either not eligible for the 100,000 Genomes Project or for whom a diagnosis is required more urgently to aid clinical management, prenatal testing or pre-implantation genetic diagnosis. For lethal fetal disorders where there is limited fetal DNA available, we recommend the parental exome sequencing strategy (Ellard et al 2015 PMID 24961629). This is a novel strategy of performing exome sequencing of parental DNA to identify potential causative heterozygous mutations, followed by co-segregation analysis by Sanger sequencing using fetal DNA. 

Please contact Dr Júlia Baptista (julia.baptista@nhs.net or 01392 408247) and Dr Emma Baple (ebaple@nhs.net or 01392 405749) to discuss the most appropriate testing strategy for your patient. For some families a bespoke virtual gene panel test is the only option due to DNA sample availability or pedigree structure. Information on read depth coverage and selection of genes for a virtual panel analysis will be provided.

For details of exome sequencing prices and testing strategies please see the Exome Request Form here.

We generally recommend “whole” exome sequencing of 23,244 genes (Agilent SureSelect v6 exome) using an inheritance based, gene agnostic approach. We have five examples where disease-causing variants were identified in a gene not included in the Agilent “clinical” or focused exome (6,110 genes). The other advantage of sequencing the “whole” exome is that the data can be re-analysed at a later date to include newly identified disease genes. Occasionally we identify variants in a gene which is good biological candidate (but not a known disease gene) and our approach is to seek additional cases via international data sharing initiatives such as GeneMatcher or DECIPHER.

Diagnostic yield and test sensitivity

The likelihood of finding a genetic diagnosis by exome sequencing depends upon the patient phenotype, i.e. the prior probability that the patient has a monogenic disorder. A successful diagnosis requires that the disease gene is known, the gene is included in the exome capture and that the mutation type is detectable by the sequencing technology. Our high diagnostic yield for families tested to date is 40% (145 diagnoses/366 families reported, 4th January 2017) with the highest diagnostic yield (44%) obtained for couples with ≥2 pregnancies affected with a lethal prenatal/neonatal disorder

We use Agilent exome capture reagents and Illumina next generation sequencing. Read depth is a key determinant of variant detection sensitivity. In our “whole” exome assay (using the Agilent v6 capture) the average percentage of coding nucleotides of targeted RefSeq exons (https://www.ncbi.nlm.nih.gov/refseq/) with ≥20 reads is 98.8%. The sensitivity for heterozygous single nucleotide variant (SNV) detection at a read depth of 20 is estimated at 99.89% (95%CI 99.88-99.91%). Small insertions and deletions (indels up to 30bp) are more difficult to detect and depends upon both the type of indel, the size and sequence context. The sensitivity for detecting heterozygous indels at a read depth of 20 is estimated at 98.6% (95%CI 97.8-99.12%). The overall sensitivity for SNV/indel detection (based on a 3:1 ratio of SNVs:indels) at 20X read depth therefore exceeds 99% (Wakeling et al 2017, submitted). This technology also enables the detection of copy number variants (CNV) using Exome Depth analysis software. The test sensitivity is unknown and dependent upon good quality DNA so may not be as reliable as MLPA; a small number of previously identified CNVs have been detected in addition to seven new diagnoses (up to July 2017) including two single exon deletions, three multi-exon deletions, one gene cluster duplication and a 106kb deletion causing split hand/split foot malformation (Ellard et al 2014 PMID 24459211).  

For bespoke virtual panel tests we use the Agilent SureSelect focused exome if it includes all the genes thought to be associated with the patient’s phenotype. The average percentage of coding nucleotides of targeted RefSeq exons with ≥20 reads for the Agilent focused exome is 97.5%. 

DNA samples required

Please send 10µg DNA extracted from a peripheral blood EDTA sample for each individual for whom exome sequencing is requested. If your laboratory routinely measures double-stranded DNA concentration (for example using a Qubit assay), then 5µg is sufficient. Saliva samples are accepted in exceptional circumstances but DNA from blood is preferred because the variable amount of bacterial DNA in saliva will reduce the amount of sequence read data for the patient.

Risk of incidental findings

Our policy is to focus on identifying disease-causing variants of direct relevance to the clinical referral. The strategies we employ result in a low prior probability of identifying mutations that predispose to other rare diseases but the possibility of incidental findings cannot be excluded and may be discussed with the referring clinician on a case-by-case basis. Trio exome sequencing will reveal possible non-paternity (or non-maternity) and this result would be discussed with the referring clinician.

Whilst there is growing evidence that many patients are keen to receive information about additional, clinically actionable findings (Middleton et al Eur J Hum Genet 2015 PMID 25920556), we await the results of studies such as the 100,000 Genomes Project (www.genomicsengland.co.uk) to understand the overall risks and benefits of receiving such information. Meanwhile the variant data from exome sequencing tests will be stored long-term in order that further analysis can be undertaken in the future.

Reporting times

Our exome sequencing test currently takes 2-10 weeks for routine referrals and 2-4 weeks for urgent referrals from receipt of suitable samples to issue of the report(s). This includes the pre-sequencing processing, next generation sequencing, data analysis and confirmation of likely pathogenic variants. The exact timings depend on factors including sample batching, scheduling of next generation sequencing runs and any instrument breakdowns as well as the time required to obtain additional information or samples for co-segregation testing. The national target for clinical exome sequencing is that >90% of reports are issued within 16 weeks (Association for Clinical Genetic Science Guidelines).

No diagnosis found

For trio or parental exome analysis where likely causative variant(s) are not identified through the inheritance based approach, we may undertake further analysis of a gene panel. The gene panels may be a bespoke design based on the phenotype using HPO terms or a previously curated gene panel (for example the Genomics England PanelApp).