Should we read the genomic sequence of all newborns?
It's a complicated question, and a new UK initiative raises a host of issues.
The UK government’s plan to introduce genomic sequencing for all newborn babies in England within the next ten years could best be characterised as bold – if by that we understand it to be expensive, of unproven value, and beset with ethical difficulties. It is possible that such a programme could bring health benefits, but the complexities are far murkier than is readily apparent from the triumphalist announcements.
Health Secretary Wes Streeting has said that “genomics presents us with the opportunity to leapfrog disease, so we’re in front of it rather than reacting to it.” It isn’t clear what this means. The makeup of our personal genome can indeed reveal something about our susceptibility to various diseases, but the connection of DNA sequence to risk is in general complicated and of limited predictive value.
And risk is not the same as prevention or cure. Streeting adds that “With the power of this new technology, patients will be able to receive personalised healthcare to prevent ill health before symptoms begin, reducing the pressure on NHS services and helping people live longer, healthier lives.” Sounds good in principle, but given that so much of the burden of disease and ill health is linked to socioeconomics, this could seem a bit like trying to improve numeracy rates by promising everyone a supercomputer.
Just about every known human trait, including susceptibility to diseases, can be correlated with differences in the composition of our genes. By studying those correlations, it has sometimes been possible to identify specific genes that appear to have a strong impact on particular diseases. For example, it has long been known that certain variants of a single gene can give rise to sickle-cell disease or cystic fibrosis. In such cases, it is already becoming possible to produce treatments and cures using gene-editing technology, as well as from the traditional approach of developing drugs to target the errant gene variant.
Those, however, are the exceptions. Most single-gene diseases are rare. And because they often run in families, people who carry the risky gene variant often know it – and some now use genetic screening during IVF to avoid passing the variant on to offspring. The NHS and Genomics England are already running a study to sequence the genomes of 100,000 babies (the Newborn Genomes Programme) to diagnose and seek treatments for such rare conditions.
In contrast, the common diseases that place the biggest strain on healthcare resources, such as heart disease, cancers, and diabetes, have very complex genomic risk profiles. Many genes are correlated with the risk of the condition, in ways that are probabilistic, not deterministic, and dependent too on lifestyles and other factors. Knowing that your genomic profile puts you in a high-risk category for, say, high blood pressure or obesity can be useful in giving you advance opportunity – in principle – to adapt your lifestyle or diet and minimize the risk. On the other hand, conventional medical testing already can and does often identify an actual rather than hypothetical cause for concern: high blood pressure, say.
The effectiveness of neonatal genome sequencing depends on having good information for interpreting the results: at present, minority communities are under-represented in genomic databases. Besides, genetic indicators even for some single-gene diseases do not guarantee that the disease will manifest – and alarmingly, there are no guidelines on the best clinical practice for children who are asymptomatic in such cases.
The Newborn Genomes Programme (NGP) already illustrates some of the complications. It is a curious mixture. On the one hand, the project will screen the genomes for more than 200 rare diseases, mostly single-gene conditions. The potential value of doing this at an early stage is clear. On the other hand, the project will, with consent, store the entire genomic sequence data for research pruposes, in anonymized form. This too can be valuable - but in a way that is essentially unrelated to any health benefits to the individuals concerned. Which makes one wonder how the project will realise its third objective: “to explore the risks and benefits of storing an individual’s entire genome [data] over their lifetime.” How exactly will they benefit if that data is never used to assess, say, risks beyond those for the rare diseases in the initial screening? Is the idea that health problems for the individual surfacing in later life might then be assessed by referring back to that personal data held in some data bank but not previously used in this way? Who then owns that data? Do the individuals have a right to see it on demand? What if they - or the parents - want to send it to a private company for genomic profiling? What happens if an individual develops a condition that might have been signalled in the genomic data if only anyone had looked? Could they sue? How secure is the data, both to hacking and to being wiped on some political whim? The NGP has clearly given some thought to such complex questions, but it’s not clear what the answers are.
There is also the question of how we should handle genetically indicated risks of conditions for which there is no known cure? If you have a genetic susceptibility to early-onset Alzheimer’s, would you want to know that from childhood? At present the NGP apparently only tells people about conditions that are “actionable”, and sets thresholds of risk for doing so, so that people are less likely to receive a false alarm. This seems reasonable, but it’s complicated. Risk assessments are typically subjective, for various reasons. Take cystic fibrosis, which can be caused by mutations to a single gene. Some of these gene variants are guaranteed to give rise to the disease - they are said to be 100% penetrant. But there are others with rather low penetrance: many people will have them and not suffer the disease, but some will. It is far from obvious how one chooses when to tell a carrier of such a variant whether to worry or not. This is not an argument to simply forgo the knowledge in the first place, but it means that the pros and cons are not obvious.
And that is one reason why that third objective - to explore the risks and benefits of genomic sequencing from birth - seems odd. Or rather, the objective itself is fine (because this is entirely the issue) - but why then roll out the screening and sequencing programme to all neonates before even having the answer? (The NGP only started recruiting participants last October, and won’t deliver its findings until 2027.) Asking this question at all is a recognition that the value of such a roll-out has yet to be evaluated.
Perhaps most importantly of all, genomic screening is pointless without the resources to follow up. A 2023 report from the Nuffield Council on Bioethics on the Newborn Genomes Programme said that “It is clear that in the current climate, NHS staff cannot undertake this additional work in the antenatal or neonatal clinics without compromising their other responsibilities.”
There is undoubtedly plenty of medical benefit still to be tapped from genetic screening and genome sequencing. Identifying the risk of rare genetic diseases at the earliest possible stage makes plenty of sense (even though, as I say, we are not currently ignorant of such risk in many cases where a condition runs in the family). And a study last year showed that, when used for children with suspected cancer, genome sequencing led to a small improvement in how the cases were managed (for example in terms of testing). But a blanket neonatal scheme will be very costly and there is no proof that it would make a difference, or indeed that it might not result in over-medicalizing of the care of healthy babies. The worry is that this becomes a fetishized technology that we use because we can rather than because we know it can make more of a difference than if the money were directed towards other aspects of healthcare.