Genetic testing: do we really want to know?

It was one of those calls you never forget. My uncle John, my mother's eldest brother and de facto head of our family, was on the phone trying to tell me he had prostate cancer. At that moment, in the winter of 2006, we did not yet fully appreciate the seriousness of his illness, but it soon became apparent he was incredibly unwell. He died within a year, at the age of 64.

Shortly after John's diagnosis his two younger brothers were tested for prostate cancer, and also found to be suffering from the disease. Alan survived for a decade, but died early last year, just a few weeks after his 59th birthday, following a short battle with oesophageal cancer. Bill, the middle son, recovered too, but had a kidney removed last month after developing a malignant tumour there.

It has become an article of faith in our family that we are afflicted with a hereditary curse. The belief hardened after we learnt there was a clear genetic link between prostate cancer and the ovarian cancer that killed my mother in 1984, nine months after I was born, when she was just 35.

When John died, thoughts of my own mortality were far from my mind; like most people in their twenties, I was going to live for ever. But as time passed, I started to wonder whether I might have inherited a genetic predisposition to the disease. It was only after moving to New York three years ago to cover healthcare for the FT that I discovered it was possible to find an answer of sorts - by taking a hereditary cancer test.

Hereditary testing is a lucrative business. Last year, Myriad Genetics, the oldest and largest name in the industry, generated $632m in revenues from its tests, including myRisk, which analyses 28 genes for defects known to increase cancer risk. In my case, the most relevant genes were BRCA1 and BRCA2. Men with harmful variations in either of these have a much higher risk of developing aggressive prostate cancer.

One in five women who develop ovarian cancer will also have a mutation in these genes. In 2013, the actress Angelina Jolie had a preventive double mastectomy after she took a Myriad test. Her result showed she had a mutation in BRCA1, which sharply increased her chances of developing breast and ovarian cancer. She subsequently had her ovaries and fallopian tubes removed.

The first step for anyone wanting to take the test is to seek the services of a genetic counsellor, a scientist trained to spot disease risk by combing through a person's family history. My counsellor, Shelly Cummings, works for Myriad, although others are attached to hospitals and academic institutions. During several phone calls, Cummings gently questioned me about every family member diagnosed with cancer. Where was the first tumour? Did it spread? Had it killed them? How long did it take?

Eventually she drew a detailed family tree. Men were squares, women circles. Those who had been diagnosed with cancer were shaded, while those who had died were crossed out with a single diagonal line. My mother's side of the family, the Hays, made for grim reading. Out of a nuclear family of seven - a mum, dad and five kids - five had been diagnosed with some kind of cancer, and only one had lived to tell the tale.

My aunt, Alma, is alone among the Hay children in never having been diagnosed with cancer. She also lost her mother, my grandmother, to the disease. She does not need to see a diagram to appreciate her loss. "It's looking at all the photos and realising that you're nearly the only person who is still alive," she says. "It's scary, and you find yourself asking why it has not ripped through other families in the way it has ours."

Alma, now 62, considered taking a hereditary cancer test and sought out a genetic counsellor last year, shortly after Alan died. But she didn't go through with it. "I had a panic, and part of me still thinks I should do it, but I've kind of settled into the way I was before. If I get it, I get it. There's a part of me that doesn't want to know. It could just be bad luck."

***

Hereditary cancer tests must be ordered by a medical professional, so I was referred to Julia Smith, a no-nonsense oncologist of more than 30 years who practises in the NYU Langone Medical Center. I felt like an imposter in her waiting room, sitting beside patients who were very sick, but after she looked over the family tree, she offered a medical opinion that even I could understand: "Wow."

Dr Smith drew a few vials of blood and dropped them in a padded envelope, which was couriered to Myriad's headquarters in Salt Lake City, which has become a capital of genetic science thanks to the Mormon belief in tracing one's lineage in order to baptise the dead. Thousands of myRisk test samples arrive at the company's laboratory each day, where they receive a barcode before entering a robotic production line with 856 steps: the DNA is extracted and then put through a huge sequencing machine, which looks not unlike the IBM supercomputers of yore.

The company analyses each of the 28 cancer-related genes to arrive at a raw genetic sequence, which it then compares to a "normal" gene. Everyone's sequence will differ in some way from the reference gene, but Myriad puts each variant into one of three categories: "deleterious", which means the mutation has been proven to increase a person's chances of developing cancer; "benign", in which case the mutation has been proven not to increase cancer risk; or "uncertain", which means the jury is still out.

Johnathan Lancaster, a Welsh-born oncologist who is Myriad's chief medical officer, describes this classification stage as "artistry", and says it is the step with the greatest opportunity for error. "Any of us could go out and buy a sequencer now, and we could within a week probably figure out how to set it up and get it running so that when we put DNA in we get a DNA sequence out." Making a definitive call on whether a particular mutation is risky or not is much harder, he says.

Dr Lancaster likens a person's genetic sequence to a long book and describes the body as a DNA "reading machine". Each "word" sends a signal to the body to produce a specific protein it needs to function healthily. But the problem is that most of us have genes that are riddled with typos and spelling errors; depending on the severity, these can lead to a multitude of diseases, including cancer.

A benign or harmless variant would be a misspelling like "toal" instead of "tool", which would be easily understood in the context of the surrounding words. "The DNA reading machine may stutter, slow down for a moment, but ultimately it produces the protein," explains Dr Lancaster. The problem arises when the misspelling or variation is more jarring, when what should be "tool" is spelt "toql", for example. In this instance, "the reading machine may not be able to get past that, so the protein doesn't get produced properly", he says. Such reading errors can prove deadly.

***

Controversy has followed the rise of genetic testing. When Myriad's founder, Mark Skolnick, discovered the BRCA1 gene sequence in 1994, commentators confidently predicted a new wave of eugenics by way of the designer baby. Others decried Myriad's decision to patent the BRCA genes, giving it exclusive rights to test for defects and handing it an effective monopoly until 2013, when the company's long fight to defend the intellectual property ended with a Supreme Court defeat.

The debate will become even more acrimonious in the years ahead. It is 16 years since the human genome was decoded but it is only now that the project is starting to unlock important scientific discoveries, including genetic defects responsible for cancers and other diseases, and new precision medicines to target them. Governments are engaged in a digital gold rush to build huge repositories of genetic information, as are academic institutions, often backed by large drugmakers and the computing power of technology companies such as Google. The hope is to crunch the data to spot biological clues or "biomarkers" that might be predictive of illness.

Grail, a Silicon Valley group, has raised $1.1bn from investors including Bill Gates and Jeff Bezos with a mission to develop genetic sequencing that can detect tiny pieces of tumour DNA long before they become terminal. "The ultimate goal is a blood test for early cancer detection," says Pedram Razavi, an oncologist at Memorial Sloan Kettering Cancer Center (MSKCC) in New York, who last month presented a well-received clinical trial of Grail's technology.

If efforts like Grail's are ultimately successful then it is not inconceivable that the general population could one day be routinely screened for cancer. But to what end? So far, the ability to detect illness is outpacing the discovery of new cures and treatments, pointing to a perturbing future where we know we are going to die early, but are powerless to do anything about it.

In her annual report this week, Britain's chief medical officer Sally Davies pledged to "spread the genetics dream to as many people as possible", urging doctors and managers to make wider use of revolutionary genetic techniques. But, as was clear when I interviewed her late last year, she frets that the wide adoption of genetic testing of healthy people will lead to a generation of the "worried well". Asymptomatic patients, armed with 30 pages of complex data, could bombard their GP, demanding they be tested speculatively - and at great expense - for scores of illnesses, she tells me. Such a phenomenon would further threaten the already straitened finances of the National Health Service.

Dr Davies also warns of information overload: patients will uncover countless genetic defects, but doctors will struggle to tell them what they mean. "I wouldn't be able to cope with it, and I'm pretty genome literate," she says. "We are going to have to explain to the public that there are cowboys out there giving you data that they don't understand, and that we as doctors won't be able to explain."

Then there is the risk that some patients will use their results as an excuse to live an unhealthy life, says Dr Davies. That would threaten to undermine years of public health messages about the dangers of alcohol, smoking, and high-calorie foods.

***

Little more than a week after my first visit, Dr Smith's office contacted me to arrange a consultation. The front page of my test report was emblazoned with a green header that read: "Result, negative. No clinically significant mutation identified." Dr Smith was quick to point out that this was not an all-clear. "It does not mean you can't get cancer, just that you do not have any of the hereditary cancer syndromes that we know of right now," she told me. "You should be followed more closely because there is, in someone with your family history, an increased chance of a genetic link that we don't yet know about." She recommended that I start screening for prostate cancer with immediate effect.

It turns out that even if I had tested positive for BRCA mutations, the recommendation - early screening - would have been exactly the same. Which begs the question, why take the test? Perhaps because if I had tested positive, I would have been able to alert Alma and my cousins to urge them to find out if they carried the defect, too. Those with children - or thinking about having them - might also want to find out if they were at risk. And knowing whether you have a genetic mutation can help your doctor decide what treatment options to pursue if you are diagnosed with cancer in the future.

But there are many oncologists who say I should not have bothered. "These genes might have been causative of your history, but even if you'd had a mutation, there is not in truth a lot you could do differently," says Kenneth Offit, chief of the genetics service at MSKCC.

Offit is especially critical of the Myriad test I took - a "multi-gene panel" that searches for a large number of defects, linked not only to prostate cancer. He prefers what he describes as an "à la carte" approach, where doctors only test for those individual defects that directly match up with a person's family history.

Perhaps the most controversial gene that Myriad tests for is CDH1. People with mutations in this have sharply higher risk of developing gastric cancer: roughly 70 per cent, versus 0.6 per cent in the general population. Even if a patient does not have any relatives with the illness, some oncologists will recommend they have their entire stomach removed as a preventive measure - a life-changing operation that involves adjoining the oesophagus directly to the small intestine. Other doctors believe this to be an entirely unwarranted overreaction.

The multi-gene panel is a mere drop in the ocean compared to whole genome sequencing. This procedure tests roughly 20,000 genes and can unearth defects associated with a vast array of illnesses. If analysed correctly, the results could predict whether someone develops a multitude of illnesses. The price has plummeted in recent years, from $10m a decade ago to little more than $1,000 today. Earlier this year, Illumina, which makes sequencing machines, said it reckoned it could cut the cost even further to just $100, making it affordable to almost everyone living in a developed country.

In April, 23andMe became the first company to win regulatory approval in the US to sell a direct-to-consumer genetic test that can tell people whether they are predisposed to 10 conditions, including Parkinson's and Alzheimer's. The company, co-founded by Anne Wojcicki, the ex-wife of Google founder Sergey Brin, charges $199 for the kit, which can be ordered online.

All a person has to do is spit into a tube and put it in the post. But given that there are no treatments that can delay or reverse Alzheimer's, what is the point of knowing - especially when such knowledge might lead to a lifetime of sadness and pain? What about blissful ignorance?

Dr Offit says there is a difference between "medical benefit" and "personal utility". There might not be a medicine for Alzheimer's, but that does not mean people cannot put money aside to be cared for when the time comes, or choose to take out long-term insurance, he says. (I doubt, though, whether insurance companies will be tripping over themselves to sell policies to people who are all but certain to become sick.)

Yet the biggest limitation of the test I took - and of all genetic testing today - is that the picture it paints is still so incomplete. Genetic science is a bit like our understanding of outer space: we know so much more than we used to, but still only a fraction of what is actually out there. Many scientists believe that an accumulation of genetic variations currently classed as harmless or benign could in fact be the leading cause of disease. All of those supposedly harmless reading errors, those "toals" instead of "tools", might add up to hurt us in the end.

I have not encountered an oncologist who thinks the unusually high incidence of cancer in my family is a fluke, or that my negative result in some way lets me off the hook. Instead they all suspect that there is some kind of genetic link that has not yet been discovered. "You cannot escape a family history like yours," Dr Offit warns.

But I now realise that I was never really trying to find out if I was going to get sick. After all of this, I am firmly in the "what's the point of knowing?" camp. My question was a different one: why did cancer all but wipe out an entire family? And why was it mine? For all the miracles of modern genetics, I have yet to find an answer.

David Crow is the FT's senior US business correspondent

Illustration by Bill Butcher

Copyright The Financial Times Limited 2017