Gonadal mosaicism and non‐invasive prenatal diagnosis for ‘reassurance’ in sporadic paternal age effect (PAE) disorders

Non-invasive prenatal diagnosis (NIPD) holds great promise to increase the options for women seeking prenatal testing, as it combines the benefits of earlier sampling in pregnancy with absence of procedure-related risk to the fetus. But all new procedures have costs and, in an article published in Prenatal Diagnosis last year, Verhoef et al. undertook a health economic evaluation of NIPD in the UK. They noted that ‘In [a] cash-constrained publicly funded health care system such as the English NHS, decisions about the broad programme of NIPD tests are likely to be largely dependent upon their relative costs’ and advocated that ‘careful audit of any NIPD service should be put in place from the outset’. Here, we wish to highlight that in prenatal testing for the (remote) possibility of sibling recurrence of so-called paternal age effect (PAE) disorders (such as achondroplasia, thanatophoric dysplasia and Apert syndrome), such NIPD (estimated cost of £550 per procedure) is already taking place (including in the authors’ own institution), without the health economic case having been clearly assessed. We believe that if such an assessment was undertaken, this particular use of NIPD may be difficult to justify at current prices. To explain and clarify our point of view, we note that there are three different contexts in which NIPD may take place because an increased risk of a dominant genetic disorder is suspected in the fetus: (1) one of the parents is themselves affected (risk 50%); (2) fetal abnormality is detected (risk depends on positive predictive value of test/findings); and (3) a sibling of the fetus has an apparently sporadic dominant mutation (risk is related to the possibility of gonadal mosaicism in one of the parents). Our concerns relate to this third scenario: Whilst studies of gonadal mosaicism suggest an overall risk of 1–2% for point mutations and up to 4% for chromosomal rearrangements (cost £13 750–£55 000 per case detected, economically justified), in the specific context of PAE disorders, the risk is likely to be at least an order of magnitude lower, so the economic argument is much more finely balanced. What do we mean by a ‘PAE disorder’? We previously highlighted that a small number of dominant genetic conditions, most commonly caused by mutations in the genes FGFR2, FGFR3, HRAS, KRAS, PTPN11, and RET, show a striking combination of epidemiological and molecular features that distinguish them from the bulk of disease-causing mutations. Briefly summarised, these features comprise (1) a very high apparent germ line mutation rate, (2) extreme bias towards a paternal origin of mutations, (3) markedly elevated age of the healthy father (this is the PAE by which these disorders are collectively known) and (4) the causative mutations confer gain-of-function to the encoded proteins, which are involved in signalling through the growth factor receptor-RASmitogen-activated protein kinase (MAPK) pathway. An alternative collective term is ‘RAMP’ (recurrent, autosomal dominant, male-biased, PAE). We have demonstrated that PAE disorders arise in testes by clonal expansion of rare mutations along segments of seminiferous tubules, through a mechanism termed selfish spermatogonial selection. This process occurs during adulthood; hence, despite positive selection, the mutations never attain the levels in sperm associated with classical gonadal mosaicism, which arises during early embryonic development. For example, in a total of 909 published measurements in sperm of four different PAE mutations causing Apert syndrome, thanatophoric dysplasia or Costello syndrome, the highest specific mutation level measured in any sample was 1 in 1380. In the context of NIPD, the unique characteristics of PAE disorders make them particularly suited to its use and (potentially) misuse. Indeed, stimulated by the knowledge that PAE mutations are the most common de novo germ line mutations in the human genome, have a narrow mutation spectrum and originate nearly exclusively from the healthy father, specific assays for several of these disorders were amongst the first to be reported (for achondroplasia/FGFR3, thanatophoric dysplasia/FGFR3 and Apert syndrome/ FGFR2). Based on scrutiny of the indications for NIPDpresented in these reports as well as in the article by Verhoef et al., it is apparent that in many instances, NIPD has been performed to detect recurrence of a de novo mutation (indication 3 earlier), rather than because of 50% prior risk or fetal abnormality.