Newborn Genome Sequencing
Newborn Genome Sequencing
Whole-genome sequencing of babies at birth, replacing or extending the heel-prick blood-spot screen. Two goals ride together in nearly every programme: screening (catch treatable genetic disease early) and biobanking (store the genome for future research and future re-analysis).
Programmes
| Programme | Where | Scale / scope |
|---|---|---|
| Generation Study | England | 100,000 babies; ~200 childhood-onset conditions |
| BabySeq2 | United States | ~1,000 genes, including adult-onset cancer risk (breast, ovarian) |
| Others | Australia, parts of Europe | trials to decide whether to offer universally |
Baseline: conventional blood-spot screening covers a handful of conditions; the broadest national panel (Italy) reaches ~50. DNA lifts that to hundreds and adds variants that are silent at birth.
What the early evidence supports
- Early detection changes outcomes for some conditions: the Generation Study's retinoblastoma case (diagnosed and treated promptly, April 2025) is the exemplar.
- Stored genomes have produced research findings: Britain's 100,000 Genomes Project (rare-disease patients and families) let Oxford researchers identify variants behind a childhood neurodevelopmental disorder; stored data can later be re-read for drug side-effects, treatment response and therapies that did not exist at sequencing.
The penetrance problem
Penetrance is the proportion of carriers of a disease-causing variant who actually develop the disease. Estimates were largely built backwards from already-ill patients, which inflates them:
- rb1 (retinoblastoma): early estimates >90%; a December preprint (Caroline Wright, Exeter) found fewer than a third of adult carriers of risky variants had ever had the cancer.
- ret (medullary thyroid cancer): new work also points to much lower penetrance; the standard follow-up is prophylactic thyroidectomy and lifelong hormone replacement, so overestimation translates directly into surgical harm.
Consequences of screening hundreds of low- or middling-penetrance genes:
- Patients-in-waiting. Children flagged but never ill. Cystic fibrosis already has a label for it, CFSPID (screen positive, inconclusive diagnosis), and parents of such children often question the value of the result.
- Harmful follow-up. Where the response is surgery or drugs rather than monitoring, a false-positive-by-penetrance is a treated-but-never-sick child.
Consent and data
- Most programmes bundle screening consent with indefinite research storage ("you can't take part in one without taking part in the other"); geneticists (Lucassen, Oxford; Friedman, UBC) call it mission creep, since screening is justified by benefit to the child and research by uncertain benefit to others, and the subject cannot consent to either.
- Breach precedent: 23andMe, 2023, 6.9m customers' genetic and personal data stolen.
- Downstream uses: some US insurers may already use genetic-test results in pricing (largely banned in Britain and Australia); blackmail (paternity) is a named risk.
Open questions
- Which variants to report: a treatable-and-high-penetrance-only rule reduces harm but conflicts with parents who want everything.
- Whether benefit outweighs cost at population scale once anxiety, over-treatment and false reassurance are counted; the current trials exist to answer this.
- Whether screening and biobank consent can be separated without collapsing recruitment.
Related
- Polygenic Embryo Screening — genetic prediction before implantation; shares the lesson that risk estimates are not destiny.
- Should Every Baby's DNA Be Sequenced (Economist) — source.
Research notes, not medical advice. Findings are reported with the population, dose and duration they were measured in; whether they transfer to you is a question for a clinician.