ESPE Yearbook of Paediatric Endocrinology

medRxiv preprint, September 24, 2024. https://doi.org/10.1101/2024.09.23.24314008

In Brief: The authors analysed genetic data from 2 population-scale studies, the UK Biobank and All of Us (USA), to identify reasons for the low penetrance of loss-of-function variants (LoF) on disease risks. While study biases are part of the answer, improved prediction of LoF variant pathogenicity also increases the observed impacts on disease risks.

Comment: There is a major discrepancy between the genetic findings from rare disorders and their families vs those of the same variants in large population-based studies. In the former, these variants invariably have ‘complete penetrance’ (all carriers exhibit evidence of the relevant disease). But in the latter, levels of penetrance are very often lower than 50%. This discrepancy is important with increasing use of genetic screening in (yet) unaffected individuals, as well as the interpretation of incidental findings on genetic testing of patients.

Some of this is due to the quality of data collected by huge research ‘biobanks’, for example which may not include relevant phenotypes or rely on routine clinical data and self-reports. However, another reason identified here is we may be too lenient about which LoF variants are classified as pathogenic and therefore reported in clinical test results. Stop-gain, frameshift and splice change variants are all grouped as LoF; we should recognise that the former group have more consistent and larger disruptions on gene function. Even among stop-gain variants, those that result in larger numbers of amino acids lost will have larger impacts on disease risk.

The next time you get a clinical test report of a pathogenic variant, think to ask how pathogenic it is.