NEW YORK – Researchers have diagnosed thousands of children with severe developmental conditions in the UK and Ireland using a combination of exome sequencing, microarrays, bioinformatic tools, systematic phenotyping, and clinical profiles.
"We employed multiple different methods and analyses to find new diagnoses and novel disease genes in our cohort, as well as both a clinician-led and computational approach to variant classification," first author Caroline Wright, a genomic medicine researcher at the University of Exeter, said in an email.
As part of the Deciphering Developmental Disorders (DDD) study, launched in 2011, researchers at Exeter, the Wellcome Sanger Institute, and other centers turned to exome sequencing, microarray profiling, standardized phenotype assessment, and a variant annotation pipeline to search for suspicious genetic features in 13,449 individuals with rare developmental disorders, most of them infants and children, who were assessed at two dozen regional genetic service provider sites. The majority of patients were recruited along with their parents, enabling family trio analyses.
"[T]he DDD study is one of the largest translational genomics studies to date … and involved unprecedented nationwide recruitment by hundreds of clinicians across 24 sites in the UK and Ireland," Wright explained.
The study, which reported preliminary results almost 10 years ago, focused on cases with severe phenotypes that were suspected of stemming from changes to a single gene. All told, the team tracked down diagnostic variants in more than 5,500 children, or roughly 41 percent of those profiled, unearthing pathogenic de novo variants in 76 percent of the diagnoses. They reported their results in the New England Journal of Medicine on Wednesday.
"We developed an iterative variant analysis pipeline and reported candidate variants to clinical teams for validation and diagnostic interpretation to inform communication with families," the authors wrote, noting that they used multiple regression analyses to consider a wide range of factors that may influence effective sequencing-based diagnoses.
Diagnoses were less likely for children who were born very prematurely — between 22 and 27 weeks of gestation — and those born to mothers with diabetes, African ancestry, or documented antiepileptic medication use during pregnancy.
There was an uptick in diagnoses and suspicious variants in cases with accompanying parental sequence data.
The team reported that parent-child sequencing led to one potential diagnostic variant per proband, on average, while singleton sequencing on the patient alone led to an average of 2.5 variants per person.
The ability to reach diagnoses also relied on collaborations and data sharing across many centers, the researchers explained, since each condition is rare, despite the large number of families affected by such conditions overall.
"For some diagnoses, it was only through sharing data with international colleagues that it was possible to make a diagnosis," co-senior author Matthew Hurles, a human genetics researcher with the University of Cambridge and incoming director at the Wellcome Sanger Institute, said in a statement.
In a corresponding NEJM editorial, Baylor College of Medicine researchers James Lupski and Jennifer Posey commented on the impressive size of the DDD study and noted that the results point to the potential for boosting diagnostic yield through genomic strategies and detailed clinical phenotyping.
"Although much work remains to be done in improving genomic diagnostics and in understanding disease biology, it is clear that clinical genomics in medical practice is making a difference for patients with rare genetic disorders and their families and physicians," they wrote.