September 22, 2021

I, Science

The science magazine of Imperial College

Should we all be sequencing our own genomes?
It’s often said that “it’s what’s inside that counts”. From a medic’s point of view, past the external stigmata of disease, this is widely true. For a geneticist it is sacred.

Genomes constitute an entire instruction manual for building an organism. They contain hidden heirlooms from our ancestors, be they early bacteria, Neanderthals or our mums and dads. They are the cause of an individual’s decay but a species’ redemption (Steve Jones).

Ten years ago it cost nearly £2 billion and ten years to sequence one genome … now it is soon to be less than £1000 and a matter of weeks. One company, 23andme, offers a £160 mail-order kit for a saliva sample, which can be returned for analysis of over 100 inherited conditions. The realisation that now every person could have their genome writ is causing considerable stir. The geneticist Craig Venter announced it “more important than walking on the moon – nothing much happened after walking on the moon”.

It seems a wonderful thought – and for me personally – that I should be able to see my mortal alphabet flash across a screen and tell me what diseases I might contract. An end is inevitable, but things I can modify to minimise the burden of disease appeal greatly. It’s an investment. You put savings away for a rainy day, you work hard at school to germinate your dream career, I leave the car at home when my hilly road is icy. Life is a perpetuating course in risk management.

There are very tangible benefits to screening genomes as well
Firstly inherited disorders are very common – 1 in 40 children born in the UK have some sort of symptomatic genetic condition, and congenital disease represents a third of childhood hospital admission. Identifying the risk of conceiving children with devastating conditions is an ethically controversial but very real and practical side of managing healthcare. For example some populations are at high risk of certain blood disorders, like the Ashkenazi Jewish population. For some years Jewish churches offered certain boons to Jewish couples who were genetically tested before marriage. In the acute care setting, a diagnosis of such disorders is critical to managing ill patients in terms of blood transfusion and oxygen therapy.

In some instances identifying the gene responsible for a disease has highlighted the disease pathway itself – allowing far more effective treatment. Haemophilia, a bleeding condition that runs famously in the old European Royal families, used to rely on transfusion only, or replacement of clotting factors themselves. Both are at high risk of transferring infected products – there was an HIV scandal last decade – and are also known to induce serious immune reactions in their recipients. In December last year the first gene therapy for a well-known disease was published from here in the UK. A virus was used to introduce the gene for Factor IX into liver cells so they produced it endogenously. It worked in 4 of 6 patients, and some are still producing their own factor IX 22 months later – in the world of gene therapy this is a flying start; brakes have been on since several tragic trials last decade where patients were killed. The viral administration costs £20,000 and although not yet clear whether this will need repeating, it is superior to current treatment. Currently, £20,000 injections of clotting factor are required at least one a month at a lifetime cost of £1.2million for one individual.

Another invaluable tool is the ability to design genetically-tailored drug therapy. Drugs have historically been a trial and error practice. Patients do not all show the same response to the same doses of treatment. Lithium, which has a multitude of applications, can be life-saving in one patient, utterly deadly in another. A large percentage of drugs must be chemically altered by the liver in order to become active, or to be excreted from the body. This involves a complex system of reactions, rates and preferences of which are written by a minimum of 30 different genes. Looking at these gene combinations can help determine what an optimal dose of drug will be. This has been used in clinic for the drug Warfarin, used to reduce stroke. In some people the drug is broken down too rapidly, effectively giving them a smaller dose, whilst in others it is degraded slowly, heightening the risk of bleeding. One US study claimed they could save over a $1 billion this way.

There are hundreds of equivalent examples, and many more when it comes to looking at risk factors for diseases like stroke, cancer, heart disease, diabetes and aneurysms.

So why wouldn’t we screen everyone?
It is a well-known fact that diagnostic tests create anxiety. If you tell a patient you are sending a test off to look for cancer, they will believe and fret they have cancer until the test comes back negative.

Recent examples in the media include the PIP breast implant scandal. Even though there is no solid evidence to suggest they have adverse effects, the NHS ruled that the psychological trauma of not having them removed outweighed the adverse risks of undergoing general anaesthetic (which is actually better avoided).

Anxiety creates a series of stress responses that are quite severe for both your mental and physical health, and in itself represents a risk for a multitude of diseases (there are even genes ‘for anxiety’ though the benefit of knowing this is negligible – an anxious person will only be made more anxious by the knowledge they have a predisposition to anxiety, whilst low-anxiety genes probably belong to the sort of person who wouldn’t have a genetic test in the first place). Dread has also been shown to lead to terrible decision-making.

Thus genetic testing is undertaken very seriously here in the UK, with every test and outcome being fully explained beforehand and counselled after.

The entire situation is made more complex by the observation that genes aren’t commandments – very few genes convey a 100% risk, and a lot of the time we don’t know whether genes convey a 1-99% risk, due to other gene interactions, and the effect of the environment.

Imagine the below on a results sheet:
1) This gene can cause cancer.
2) This gene is associated with cancer.
3) We don’t know anything about this gene.
4) This gene is a rare mutation.
5) You have cancer.

1) You are worried, you think you have cancer. You might get anxious and depressed, or worried for your children. You are not listening to ‘can cause’. Can could be in 1% of cases, or 100%. Hard to know.

2) Associated is a loaded word. Sunrise is associated with my alarm going off, that doesn’t mean the two are related. Sunrise is also associated with the earth rotating on its axis. These are related. Requires further rounds of tests and drawn out anxiety.

3) Unhelpful. You are probably still more concerned about cancer than if no-one had ever offered genetic testing.

4) Mutations have a bad reputation but they are in fact the reason humans are here. The majority of mutations are neutral, given they are more likely to occur in regions of DNA that do not actively write useful proteins. Some are bad, very few are beneficial (though from our point of view, those that changed between bacteria and humans, were). The issue with large-scale genetic screening is that no two people have the same genome. There will be thousands of single point mutations that have no effect on disease, but will be picked up as ‘variants’ on screening. Below are examples of neutral and detrimental “mutations”.

The red fox jumped over the dog
The crimson fox jumped over the dog

The above sentences are academic divergents, but don’t change the essential point.

The red fox jumped on the dog

Now the sentence has changed its meaning,

The red fox zebra over the dog

Now the sentence makes no sense, and any sensible editor would fix or delete it. In this analogy the words are pieces of DNA, the sentences are genes. Thus mutations, per se, are hard to interpret without broader context.

5) It’s rare a genetic test can confirm cancer without a clinical picture – this is a diagnosis delivered by clinicians only. You will have been told by an informed, empathetic individual, with all the right treatments at their fingertips, rather than a cold piece of paper.

Genes convey a relative, but not absolute, risk of illness. I may have genes that predispose me to lung cancer, but until I smoke they are unlikely to affect me. Every single one of us carries at least one copy of a dangerous gene. I would not want a print-out of all the rare but serious risks in my life i.e. plane crashes, objects dropping on my head off balconies, being struck by lightning … it’s depressing. The same can be said for my genome. Unfortunately we’re still fairly poor at deciphering just what risk a gene conveys.

Humans are a woven mosaic of gene-environment interactions. Your DNA is largely the same whether you are an egg or a 90 year old man … a lifetime’s exposure to the world helps to shape what you become … and it’s not laboratory scientists who have the complete mosaic in front of them, it’s doctors. I believe gene testing can supplement their work, but still believe whole genome screening is inadvisable, and we should stick to specific screens (such as BRCA2 for breast cancer).

Margaret Anderson said “there is a danger in knowing too much. It is a boon to people who lack deep feelings”. Like insurance companies …

Images: flickr | shaury and dullhunk