Presented at the Neonatal Society 2018 Autumn Meeting.
French CE1, Delon I2, Dolling H1, Sanchis-Juan A1, Austin T2, Branco RG1,2, Rowitch DH1, Raymond FL1,2
1 School of Clinical Medicine, University of Cambridge, UK
2 Cambridge University Hospitals NHS Foundation Trust, Cambridge Biomedical Campus, Hills Road, Cambridge, UK
Background: A significant number of patients admitted to NICU/PICU) can present with undiagnosed genetic conditions. There is therefore a need for rapid but accurate molecular diagnosis to inform decision. The objectives of this study were to establish a whole genome sequencing (WGS) analysis platform within the UK NHS that delivers clinically useful, affordable and timely early diagnoses,and to investigate the phenotype and genotype correlations in very young children.
Methods: Trio WGS analysis was performed in a prospective cohort of families recruited from NICU/PICU over an 18 month period. A hybrid pipeline between research and the UK NHS was developed to deliver pertinent pathogenic findings within 2-3 weeks of recruitment. All families were recruited following signed informed consent. A minimum of 500 μL of fresh blood in EDTA was requested in order to yield a minimum of 1 μg of DNA for WGS.
Results: In NICU, recruitment was broad, excluding those with a known genetic diagnosis and preterm neonates without additional features. 49% of families approached consented to the study and of these 90% provided a two-parent and child trio sample. Recruitment was declined by 15% of families and the remainder were undecided by the time of discharge. To date, 195 families have been sequenced and analysed, including 106 from the NICU. Sample receipt to diagnosis takes 2-3 weeks. Overall, a likely diagnosis was reported in 20% of cases and the diagnostic rate in the NICU was 13%. Diagnoses included encephalopathies, myopathies, skeletal dysplasias and rare syndromes. Neonates with clear dysmorphology (e.g. hypertelorism, micrognathia) were more likely to yield a diagnosis (~50%) but, in general, it was difficult to phenotypically distinguish those with and without a genetic aetiology. For the majority, the phenotype presenting in the N/PICU was a poor predictor for the gene abnormality that was detected because it was often atypical for the known disease-causing gene. A genotype driven approach of the whole genome data ensured that new genes with fewer documented phenotypes were considered equally with genes that have well-characterized phenotypes. The most common clinical impact of the diagnosis in the N/PICU was to direct or initiate specialist care pathways for improved management and early intervention. In 13%, the diagnosis affected acute clinical management decisions and modified treatment. For 33% the genetic diagnosis had significant recurrence risks for subsequent pregnancies. For a few families where the infant died at or shortly after birth, the genetic diagnosis provided sufficient explanation for the cause of death and for others provided evidence to support end of life care decisions.
Conclusion: The results demonstrate the new challenges of early diagnosis whilst establishing the feasibility and utility of whole genome sequencing in improving the management of rare disorders in the intensive care setting where the phenotypic spectrum of children presenting early in life with a genetic disease is still being discovered.
Corresponding author: email@example.com