It is likely parents of children are already receiving results indicating that their child has loss-of-function variant in
ELP1 or
GPR161. Indeed, in England neonates can now undergo genome sequencing at birth without any obvious symptoms and parents could receive results for an incidental pathogenic variant in
ELP1 or
GPR161 as genes associated with childhood malignancy. The present study will provide counsellors with sufficient risk information to provide an accurate risk assessment and recommendations. For instance, parents of neonates with a
GPR161 variant can be reassured that absolute risks of medulloblastoma are very small and the increased risk may disappear after age 4 years (this will be reassuring if results are given for a 5year old) similar to
SUFU and
PTCH1 [
3,
5,
6]. Whereas, parents of neonates with an
ELP1 loss-of-function variant can be told the risks are higher but not sufficient for MRI screening [
3,
6] and that the increased risk lasts well beyond 7years of age. The risks for siblings of medulloblastoma cases who are heterozygotes may still justify closer monitoring especially for
ELP1 as the family may carry additional genetic factors that predispose to medulloblastoma. Without this granular information on the true implications of this finding these parents may become distressed and concerned about the requirements for clinical screening. The implied likelihood of developing medulloblastoma for both
ELP1 and particularly
GPR161 are overall quite low and at or well below that of
PTCH1 for which no screening in childhood for medulloblastoma is recommended [
3,
6]. The suggestive features for a possible diagnosis of Gorlin syndrome in the index
GPR161 patient was not backed up by information on the father, brother and two nephews who were also heterozygous [
2]. The population frequency of both
ELP1 and
GPR161 of close to 1 in 1,000 are also far too high to account for the small amount of missing heritability in a condition with a birth incidence of only 1 in 14,500 [
7,
8]. We did not identify
GPR161 in 27
PTCH1/SUFU negative Gorlin syndrome families which represent all the 27/86-(31.4%) unexplained by known genes (59 were due to
PTCH1/SUFU) meaning an unexplained population frequency far lower at 0.0021% (1 in 46,400). Given the absence of clinical data and the even higher population frequency of
ELP1 we have not assessed this gene in our Gorlin population.
ELP1 and
GPR161 therefore join
PTCH2 as potential candidate genes for Gorlin syndrome that can be dismissed on their population frequencies and, for
PTCH2 and
GPR161, their absence in Gorlin syndrome kindreds [
8].
We do not have an explanation for the apparent higher incidence of childhood medulloblastoma in the USA compared to the UK and this seems unlikely to be linked to an increased frequency of the known predisposition genes.
In examining gnomAD, we found a ClinVar entry for a putative likely pathogenic
GPR161 missense variant, c.56T > A,p.Leu19Gln;NM_001267609;chr1:g.168,074,093 A > T[hg19] (this variant is annotated as c.-5T > A; 5’-UTR in the MANE select transcript:NM_001375883.1). The variant had been reported in homozygous form in a Turkish family as causing Pituitary Stalk Interruption Syndrome [
9]. Although this quite clear syndromic diagnosis tracked with zygosity in a consanguineous kindred (the unaffected heterozygote parents had 4 further unaffected children who were not homozygous) no population frequency for this variant was supplied. We have now assessed this in gnomAD and found that it is present 18 times in South Asians in homozygous form in 15,296 individuals (1 in 850 individuals) with a further 641 being heterozygote (1 in 24). This clearly cannot be a loss-of-function variant that could cause a medulloblastoma risk (particularly as homozygous) or this would mean a much higher likelihood of medulloblastoma in South Asia and the homozygote frequency is far too high to be associated with such a rare (estimated 0.5 per million) complex condition as Pituitary Stalk Interruption Syndrome [
10].