A Consanguineous Family with a Neurodevelopmental Disorder
In a consanguineous family, first-cousin parents are unaffected and have three children: two sons with developmental delay, seizures, and low cerebrospinal fluid glucose, and one unaffected daughter. Whole-exome sequencing identifies a homozygous missense variant, c.1234G>A (p.Gly412Arg), in a gene linked to autosomal recessive neurodevelopmental disease. Population data show allele frequency 0.002 in 10,000 matched controls, with no homozygotes; in-house data include one heterozygote among 2,000 unaffected adults. In silico tools disagree: SIFT predicts deleterious, PolyPhen predicts benign; the glycine is conserved (phyloP 5.8). A minigene assay shows normal splicing, and an enzyme assay in patient fibroblasts shows 25% of wild-type activity. The parents are heterozygous. The family asks about recurrence risk and whether the variant should be reported as pathogenic.
1. Calculate the recurrence risk for the couple’s next child and the carrier frequency in the population. Explain how the population data affect interpretation.
Read the worked answer
Because the parents are first cousins and both carry the variant, the disorder is autosomal recessive. The next child’s recurrence risk is 25%; an unaffected child has a 2/3 chance of being a carrier. Population carrier frequency from q = 0.002 is about 2pq = 0.004, or 1 in 250. The absence of homozygotes among 10,000 controls is expected: q² × 10,000 = 0.04. The in-house 1/2,000 heterozygote count is based on a small sample and does not contradict a 1/250 carrier rate. Thus the population data do not argue strongly against pathogenicity, but they do mean the variant is not extremely rare.
2. Using the provided evidence, decide whether the variant should be classified as pathogenic, likely pathogenic, or uncertain. Discuss the conflicting in silico and functional data and what additional evidence would help.
Read the worked answer
The evidence is suggestive but not conclusive. Supporting: homozygosity in two affected siblings, conservation (phyloP 5.8), and 25% enzyme activity. Against: conflicting in silico predictions, normal splicing, and a 0.002 allele frequency with heterozygous adults. Under ACMG-style reasoning, the functional assay may count as moderate/strong evidence only if validated; PM2 is weak or not applicable at this frequency. I would classify it as a variant of uncertain significance unless the gene–disease match is specific and the enzyme assay is validated; likely pathogenic would require additional segregation, unrelated affected cases, or a rescue experiment. Additional evidence: mRNA/protein levels, homozygote frequency in patients, and functional rescue.