27 December 2019
In 2003, humanity completed perhaps its most ambitious voyage of exploration when scientists mapped the human genome — the three-billion-plus pairs of DNA that make us human.
The project was supposed to unlock enormous benefits to health. We would finally be able to read our own blueprints and see how our lives were likely to run. It has taken another 15 years but genomics is about to make good on that promise.
Within the next few years, experts believe, a visit to the doctor might include a quick, cheek-swab DNA test.
A few weeks later, you might receive an email listing your "genetic risk" for a range of conditions including cancer, stroke and heart attack.
Genetic risk scores will soon become a part of your health record, says Associate Professor Michael Inouye.
"It's not a question of if but when," he says."It's inevitable."
The science, it seems, has now caught up to the hype. "The genomics era is finally upon general practice," the Australian Journal of General Practice editorialised earlier this year.
But, as with all powerful technologies, it poses difficult questions.
What is polygenic testing?
The new technology is known as a polygenic risk score.
Scientists spent decades going through our genes one by one to see if they made us more likely to get sick. These genetic tests are very useful for conditions caused by an error in a single gene, such as Huntington's disease and cystic fibrosis.
Consumer DNA tests such as 23andMe (not available in Australia) also look at a small number of genes and try to gauge your risk for various conditions, although doctors warn the accuracy of their test is not always great.
But for most diseases and conditions, it turns out that single genetic variants only have a very small effect on risk. That is, the function of individual genes matter far less than the overall picture drawn by your DNA.
"You get all these small effects spread across the genome," says Dr Inouye from the Baker Heart and Diabetes Institute.
We're still not sure why this is, nor what all these small effects are caused by.
But rather than worrying about that, scientists have focused on turning this insight into a clinically useful product: a polygenic risk score.
They have used supercomputers to add together all the tiny variations across a person's entire genome to work out their risk for various conditions.
Much of the pioneering work was done in agriculture, trying to predict which cows would produce the most valuable offspring; it was a natural step to take that into human health.
How does polygenic testing work?
A polygenic test will use a small sample of your DNA — usually taken via a cheek swab. That will be sent away for testing. Your GP will receive your results in a few weeks.
The results you get back won't tell you if you will develop, say, bowel cancer. It can only tell you whether you are at high, normal or low risk.
"The most important thing to recognise is these are predictors — they are not diagnostic," says Professor Naomi Wray, a statistical geneticist at the University of Queensland.
"Even if you're in the top 1 per cent for a risk score, most people in that group still won't get the disease."
Conversely, even if you're at low risk, there is still some chance you will get a disease.
The risk expressed in the test is based solely on your genes, and does not incorporate other risk factors such as obesity, diet and smoking status.
Currently, GPs offer risk scores based on those lifestyle factors, as well as a family history of, say, heart disease. The new genetic risk scores won't spell the end for these kind of assessments; in some cases, they will be more accurate.
But the genetic tests will be especially useful for people who do not know their family history for diseases — such as if their great-grandpa really did die of a heart attack.
How will the results be delivered?
The big question still unanswered: how will you receive your polygenic score?
With the science now mature, scientists — and governments — are now busy working out how to implement the technology in healthcare. This is the final hurdle that needs to be overcome before risk scores are widely rolled out.
The obvious solution is to give everyone a full readout of their risk scores for dozens of conditions. But how do you interpret being told you're at increased risk of a disease when the actual risk remains very low?
"Generally speaking, people are not very good at understanding relative risk," says Professor Peter Visscher, a quantitative geneticist at the University of Queensland. "But you don't want to withhold information from people."
Relative risk is your increased (or decreased) chance of catching a disease relative to the average person. Polygenic risk scores cannot tell you your absolute risk — your actual chance of catching a disease.
One option may be to only give you scores for things you can do something about. Bowel cancer polygenic scores, for example, would allow people to be more vigilant about screening.
What can it show? What can't it show?
"In principle, one can make a score for many, many common diseases," says Professor Visscher.
That includes heart disease, stroke, Alzheimer's, dementia, obesity, breast cancer, ovarian cancer, prostate cancer and some auto-immune diseases.
Dr Inouye has developed a score for heart attacks.
His data shows that people with high-risk DNA are more than four times more likely to have a heart attack than people with low-risk DNA.
And some women with the "right" DNA have almost no risk of having a heart attack even if they have high blood pressure and high cholesterol.
The question is no longer whether we can predict things based on our genes — it is whether we should.
Alzheimer's, for example, has no effective treatment. "What do you do with that information when there is nothing you can do about it?" says Professor Visscher.
Other conditions, such as obesity, are influenced more by lifestyle choices than genes.
A genetic test to predict how long we are likely to live is theoretically possible, for example, "but it wouldn't be very good. We know most differences in lifespan are not based on genes — if you're hit by a bus, for example," says Professor Visscher.
Experts said they expected similar cardiovascular risk tests, along with risk tests for several cancers, to be the first things to be widely rolled out.
Do the tests work for everyone?
So far, the vast majority of polygenetic tests have been developed for people with European ancestry.
It's not clear how well they will work for, say, Indigenous Australians or Africans, who own very different DNA. Genetic Technologies' bowel cancer risk score, for example, can only be used on Caucasians.
The answer to that is simple, says Professor Wray. Get more data.
What will you do with the information?
How would you deal with being told you are at higher risk of cancer than average — even when that risk is still low overall?
And what would you do if you were told you could eat hamburgers all day and still have almost no risk of suffering a heart attack?
"This is all really wide open. And we're at the critical moment. There are all sorts of paths that could become reality," says Dr Inouye.
Professor Jon Emery, a former GP now based at the University of Melbourne, is in the middle of a trial offering genetic risk scores for bowel cancer at GPs' offices.
The scores won't eliminate the need for regular bowel cancer screening. But Professor Emery hopes they will prompt those with high scores to be extra vigilant.
But he cautions that people with low-risk DNA still need to pay attention to warning signs. "Even if you're at low risk, you're still potentially at risk. So you still need to look after your health."
Who else gets to know?
There is another other key breakthrough that makes polygenic risk scores possible: insurers have agreed to stop asking for them.
Previously, if you received a genetic test, you were legally obliged to provide the data to your life insurance company. Bad results could potentially make you uninsurable.
In June, Australia's life insurance companies agreed on a moratorium of the practice. You no longer have to disclose a bad result, but you can voluntarily disclose a good one — which could see your premium fall.
"You cannot be worse off with life insurance, you can only be better off. There is no downside," says Nick Kirwan, senior policy manager at the Financial Services Council. "Life insurers are the beneficiaries of genetic testing. We want to encourage people to take part."
Who profits from all this testing?
So far, governments have not been willing to stump up the huge sums needed for population-wide genetic screening.
That's understandable as the technology has not yet been mature enough; but it has also left the door open for private companies.
Many polygenic risk score formulas are released freely by scientists, meaning anyone can apply them to sequenced DNA.
That's already happening. Genetic Technologies, an ASX-listed company based in Fitzroy in Melbourne, plans to launch risk scores for breast and bowel cancer to the public early next year. 23andMe have their own polygenic score in the works.
If private companies do take the lead — offering branded scans in GPs' rooms, for example — that raises thorny issues.
Will the technology's benefits accrue only to those who can afford them? And what will the companies do with the extremely valuable genetic data they gather?
"Private companies will jump on this. They will want to sell every test for every disease individually and make a lot of money," says Professor Wray.
She hopes to see polygenic tests trialled through GP clinics soon. She's even been pondering whether infants should be DNA tested, with all the data stored on their health record.
"I think we'd almost forget we had our genetic data on a health record," she says. "A polygenic test is just another variable that is a little bit useful."
Stay on top of the latest findings with our newsletterSubscribe