A R T I C L E 6. It’s in your genes – maybe.
Jul. 18
th
2008
From Economist.com
Peering into your medical future is risky
IT HAS already delivered ever cheaper and more powerful computers. Now Moore’s Law – the prediction four decades ago by
Gordon Moore, one of the founders of Intel, that computer chips would roughly double in performance every 18 months or so – is
promising to turbo-charge our health care as well.
The "genome chip" – a matchbox-sized micro-array, fabricated on a slither of silicon or quartz, that can detect 1m or more
specific genetic variations in an individual’s DNA at a time – is following an even steeper price-performance curve than Mr Moore
ever imagined.
In 2003, when the first human genome was decoded, the overall cost was about $3 billion. By 2005, the cost for a similar job had
fallen to $15 m or so, thanks to speedier gene chips built by the likes of Affymetrix and Illumina, a pair of bio-tech start-ups. Another
firm, Knome of Cambridge, Massachusetts, is now offering to sequence an individual’s complete genome for only $350,000.
And it doesn’t stop there. George Church, a geneticist at Harvard University and a pioneer of gene-sequencing technology,
expects personal genomes to be available for as little as $20,000 within a few years. America’s National Institutes of Health (the NIH)
has set a goal for sequencing complete human genomes for $1.000 by 2014. The way things are going, the $1,000 genome will be
available long before then.
To make that happen, the X Prize Foundation in the United States–the same outfit that put up a $10 m reward for the first
commercial spacecraft to fly to the edge of space twice within two weeks – is offering $10 m to the first private team to decode 100
human genomes within ten days for less than $10,000 apiece. That’s the biggest medical prize ever.
The key to the whole endeavour is the rapid progress made over the past couple of years in genotyping equipment. At the heart
of the technology is a micro-array consisting of thousands of microscopic spots of different sections of DNA deposited on a glass or
silicon surface. When fed a sample of someone’s DNA, these spots detect single-letter changes known as "single nucleotide
polymorphisms" (SNPs, pronounced "snips").
SNPs are single-letter mutations within the DNA that determine how one person is different from another. As such, they are
signposts along the genome that provide clues to what version of a gene a person may have, and whether it’s been linked to some
particular disease.
Using SNP technology, researchers have been finding new disease markers weekly. We now know the SNPs for cancer of the
breast, prostate and lung as well as for obesity, diabetes, rheumatoid arthritis, Alzheimer’s, multiple sclerosis, cardiovascular disease,
deep-vein thrombosis, schizophrenia and countless more.
With the technology racing along so fast, entrepreneurs have jumped aboard the bandwagon to offer members of the public
tantalising glimpses of what their genes may have in store for them. 23andMe of Mountain View, California, and deCode Genetics of
Iceland started last November. Navigenics of Redwood Shores, California, got going in April.
The services on offer check your DNA (from a sample of saliva) to determine your risks of contracting various diseases and
conditions later in life. The tests don’t come cheap. Navigenics, which offers extensive follow-up and counselling services, charges
$2,500 for a scan, plus a $250 annual fee for scientific updates. The other two charge around $1,000 per screening.
These are not, of course, complete sequences of the customer’s genome – you have to go to Knome or places like Dr Church’s
Personal Genome Project at Harvard for that. What they do instead is scan your DNA for markers associated with specific diseases.
23andMe tracks 58 diseases and conditions, deCode offers 26, and Navigenics 18.
What you get back from the scan is a report showing the risk you run, relative to the average, of contracting the various
conditions tracked. The companies involved go to great pains to explain that this is not a diagnosis, just information about your
particular make-up.
That’s largely because firms like 23andMe and Navigenics have to walk a fine line between providing medical information
about patient’s potential health and actually performing diagnosis. In America, the Food and Drug Administration strictly regulates
diagnostic testing for disease, but has been slow to extend its oversight to the public implications of genomics.
With the federal authorities slow to react, state governments have begun to crack down. Health officials in New York issued
cease-and-desist notices to a dozen or so testing companies last April. Last week, California followed suit.
The authorities worry that gullible members of the public may react too hastily to their genomic information. They see the
unproven reliability of such tests as at best a waste of money, and at worst a danger to public health.
The fact is, no one knows for sure how many of the genetic tests being pushed for various conditions are actually useful. Some
may be misleading or worse. The potential for false positives, needless surgery and untold anxiety is enormous.
And what about those who score below average for critical conditions like cardiovascular disease or diabetes – and then abandon
healthy ways of life because they think they are immune? The potential for false negatives is no less acute.
The biggest problem is that the gene-sequencing technology is galloping ahead of the medicine for treating many of the
conditions implicated – to say nothing about the ability of the medical profession to understand how best to handle such information.
The problem is we’ve acquired a sort of blind faith in genetic testing over the past quarter century. Since 1983, when
Huntington’s disease was found to be correlated with a particular chromosome, health researchers have identified no fewer than 1,400
diseases associated with single genes. These so-called "monogenic" tests have been remarkably accurate in predicting the likelihood
of a person getting the disease in question.
But finding reliable monogenic links to disease is the easy part. Unfortunately, the "one-disease-one-gene" approach to genetic
testing applies to only 5 % of diseases. The other 95 % are each related to a subtle interplay of many different genes and other factors.