Last week’s landmark report on personalised medicine plays down potential for harm and oversells uncertain benefits.
One of the biggest challenges in medicine is how to offer hope without hype. Designing personalised treatments based on the science of the human genome will doubtless bring benefits, particularly for those with rare genetic conditions. But over-promoting the promise risks a tsunami of unwarranted diagnoses and unnecessary treatment. Without rigorous evaluation, widespread premature implementation of this costly new approach threatens human health and health system sustainability.
The influential Australian Council of Learned Academies published an upbeat report this week on the future of precision medicine, where scientists work to determine the genetic and biochemical makeup of an individual and doctors use that information to tailor personalised treatments or prevention strategies.
The report aimed to examine the opportunities and challenges posed by precision medicine in Australia yet it makes only passing reference to the potential harms of screening and diagnosing the genes of healthy people. The very real risk of widespread over-diagnosis is not mentioned once in its 200 pages.
Over-diagnosis happens when people are diagnosed with a disease that would never actually harm them. It happens because the current technology we use to diagnose – like ultrasounds or CTs scans – can detect ever-smaller “abnormalities”. Despite many of these “abnormalities” being benign, they tend to be routinely diagnosed and treated. A classic example is the over-diagnosis and over-treatment of tiny thyroid cancers, many of which would never cause any harm if left alone.
There is increasing recognition that early detection is a double-edged sword. For some people it can extend life. For others it means unnecessary treatment that’s harmful and costly. The potential side effects of treatment for thyroid cancer include vocal cord damage and life-long medication.
Evidence suggests over-diagnosis already happens across many conditions and is part of the wider problem of too much medicine: too many antibiotics, CT scans for kids and prescriptions for the elderly.
Assessing all possible genetic mutations that might cause someone to have future disease clearly carries great potential to increase over-diagnosis. Many people will be classified as sick or “at risk” of being sick, even though they’ll never develop any symptoms.
Also missing from the hype in this week’s report is the great uncertainty surrounding promised benefits. According to a recent analysis, many newly discovered so-called disease-carrying mutations may have limited or no meaningful effect on future illness. In a nutshell, there’s still huge unanswered questions around how to interpret results of genomic testing.
Most ominous is a theme in the report suggesting health regulations may need to be watered down to help facilitate this new approach of tailoring treatments to individuals. It calls for “agile regulatory conditions” that don’t inhibit implementation; it says current processes for evaluating new technology may be “too cumbersome”; and cites a global push to see regulation standards “reduced”.
In contrast, many would argue technological innovation must co-exist with rigorous evidence-based evaluation, if people are to be properly informed about risks and benefits.
The full report does have valuable information, including acknowledgement of huge costs attached to what is still largely an experimental new approach to individualised treatment: “enormous sums of money need to be spent on treating patients who are unlikely to respond, in the hope of helping the few who will”.
It also rightly urges examination of the social, legal and ethical implications of genomics and calls for community dialogue. Hopefully that will be a chance to critically debate pros and cons, rather than promote and persuade.
There will inevitably be advances from this new science – better analysis of infections, of antibiotic resistance, even a reduction in over-diagnosis are among the real possibilities. But some scientists are less sanguine. Writing in Nature, cancer specialist Vinay Prasad gave this assessment of the prospects for designing individual drug therapy using genetic mutations in tumours: “At best we may expect short-lived responses in a tiny fraction of patients”.
Not everyone sees combining genetic science with information technology as a panacea. In a piece published in the leading Journal of the American Medical Association, internationally respected Professor John Ioannidis and colleagues declared “this approach has largely failed”. The paper argued despite billions in public funding for genetics and related areas, there was little evidence of actual improvements in health. They urged research funders to support more work of clear public health importance.
Last week’s report correctly argues expansion of genomics and precision medicine will bring benefits to industries selling it. Whether most Australians and our health system will similarly win, is a lot less certain.
Dr Ray Moynihan is Senior Research Fellow and NHMRC Early Career Fellow at the Centre for Research in Evidence-Based Practice at Bond University, and part of the Wiser Healthcare Research Collaboration.
This is a version of a piece published in the Opinion Pages of the Sydney Morning Herald, Saturday 3 February 2018