Are supposed “high-risk” genetic variants less risky than we thought?

Many recurrent genetic changes linked to neurodevelopmental disorders may be far less risky than earlier research suggested, according to new research by Goh et al. published in the European Journal of Human Genetics.

Copy number variants (CNVs) are stretches of DNA that are deleted or duplicated, and some have been linked to increased risk of intellectual disability (ID). For years, published estimates suggested that certain CNVs carried a 10-40% chance of causing ID. But researchers now say these figures were inflated because they counted unrelated background risk of the condition.

To address this, the team redefined penetrance in more clinically relevant terms, as “the probability of manifesting a phenotype due to having the genetic variant,” explicitly excluding unrelated causes. They developed a new mathematical formula to calculate penetrance, drawing on data from eight major studies of children with ID or infants with developmental delay and a large control cohort from gnomAD. Using the prevalence of paediatric ID in high-income countries (1.1%) as a baseline, they recalculated penetrance for 83 recurrent CNVs. The authors note that the same mathematical approach could also be applied to other classes of genetic variants.

The results showed a dramatic downward revision. Most CNVs once thought to carry a 10-40% risk of ID are now estimated at just 1-10%. Some, such as 1q21.1 proximal duplications, 2q13 duplications, 15q11.2 duplications, and 16p12.2 duplications, were effectively recalculated to 0% penetrance for ID. In contrast, full penetrance for well recognised syndromes such as Williams-Beuren or Prader-Willi/Angelman remained unchanged.

There are potential limitations. The penetrance estimates are sensitive to the assumed background prevalence of ID and may be higher if a larger figure for this is applied. Moreover, the new calculations are limited to ID and do not account for other phenotypes that may be associated with CNVs, such as developmental delay, autism, epilepsy, and congenital malformations.

However, researchers say the new estimates will help prevent overdiagnosis, reduce unnecessary anxiety for families, and give those providing genetic counselling more accurate tools. The findings are expected to influence how labs interpret CNVs in diagnostic and prenatal settings, shifting the understanding of which genetic variants are truly risky and which may pose little or no risk.

Subscribe for the latest updates delivered straight to your inbox