Search

Rare diseases, particularly those with great phenotypic and genetic heterogeneity, remain a diagnostic challenge. Broadening the investigative approach through whole genome sequencing (WGS) is anticipated to improve diagnostic yield. Despite this, substantial numbers of patients remain undiagnosed through current 100,000 Genomes Project pipelines. In these cases, the causative mutation may not be prioritised due to lack of evidence associating the gene with a specific phenotype, or as a result of being inadvertently filtered out. A systematic approach to additional data analysis utilising both inheritance model filtering and ranking algorithm tools has been piloted in cases with no primary findings (NPF). Examples of diagnoses secured include a de novo TAB2mutation, causing a connective tissue phenotype with congenital heart disease and a paternally inherited COL7A1mutation, causing a variable penetrance skin blistering disorder, epidermolysis bullosa dystrophica. Identification of causative variants in WGS data remains challenging. Data led approaches to optimising additional review of cases with NPF have the prospect of improving both sensitivity and cost efficiency and may provide earlier diagnosis for individuals with rare disease, ending their diagnostic odyssey. Establishing a standardised approach to reanalysis will be particularly important as we approach the introduction of an NHS clinical sequencing service.

Background Great Ormond Street Hospital cares for many children with rare and complex disease. A substantial proportion of these children will have an underlying genetic cause for their condition. Making a diagnosis in such cases is increasingly achievable due to advances in genomic testing. However, such testing may not be available on the NHS and results, when accessed in the research setting, typically take many months to report. In critically ill children, cared for on the intensive care units, a timely genetic diagnosis may be beneficial on a number of levels. It may inform treatment options and prognosis. It may also reduce the need for further investigations, improving the diagnostic odyssey and potentially reducing costs. In addition, it may help the family understand recurrence risks. Methods To address this unmet need we developed a work flow to enable Rapid Paediatric Sequencing (RaPS) in critically ill children. Eligible patient-parent trios were recruited and underwent whole genome sequencing. A framework was developed to allow capture of clinical information and robust analysis of genomic data initially by investigating candidate gene lists and HPO-derived panels. If no genetic diagnosis was forthcoming, genomic analysis was broadened and more disease-causing genes were investigated. Results Twenty two patient-parent trios were recruited to the RaPS pilot project. A diagnosis was reached in 8 cases (36%). The shortest turnaround time from sample collection to return of provisional results was 5 days. A genetic diagnosis informed point of care decisions in at least 3 cases. Conclusion We report on the process and framework necessary to deliver rapid whole genome sequencing to critically ill paediatric patients and reflect on the outcomes, clinical utility and challenges of implementing such a service at GOSH.

Background The 100,000 Genomes Project is a national programme to perform whole genome sequencing (WGS) in families with an undiagnosed rare disorder or cancer. Multiple specialties from GOSH have recruited patients to the pilot study. Here we report on the outcome of the first 111 pilot cases. Methodology Total lymphocyte DNA was extracted (Chemagic Star, Chemagen) and exported to Genomics England for WGS (HiSeqX, Illumina). Prioritised variants were discussed and classified by a multidisciplinary team (MDT) of Clinical Scientists, Consultant Clinical Geneticists and the recruiting clinicians. Variants deemed to be ‘pathogenic’ or ‘likely pathogenic’ were flagged for diagnostic confirmation and clinical reporting. Results Genomics England returned results for 111 cases (probands); 95 (86%) were analysed as trios (proband and both parents). Variant interpretation has been completed for 97 cases (87%). Thirty-six percent of cases (40/111) and a total of 76 variants were discussed at an MDT meeting. Fifty-three percent of variants discussed (40/76) were classified as pathogenic or likely pathogenic, 30% (23/76) were classified as variants of unknown significance and 17% (13/76) were classified as benign or likely benign. This equated to a genetic diagnosis in 27/111 cases (25%) of which 5/111 (5%) were felt to be a partial rather than full explanation of the phenotype. Of the 97 cases where primary analysis was complete, 66/97 (68%) had no primary findings. This included cases in which variants of unknown significance have been detected. Conclusion WGS is a useful tool for rare disease diagnosis especially in patients who have undergone a fruitless diagnostic odyssey. Analysis of GOSH pilot cases has enabled us to develop infrastructure locally to support the return of main programme results. One quarter of GOSH pilot patients should expect a genetic diagnosis from the primary analysis. It is unclear whether this will be an accurate reflection of the diagnostic rate achieved in the main programme due to ongoing refinement of the recruitment and analysis pipelines.