The genomic clues to disease
Genomics has moved out of research and into patient care, finding and identifying gene changes that can help with the diagnosis and treatment of genetic diseases
CHRIS HATZIS
Eavesdrop on Experts, a podcast about stories of inspiration and insights. It’s where expert types obsess, confess and profess. I’m Chris Hatzis, let’s eavesdrop on experts changing the world - one lecture, one experiment, one interview at a time.
Discovering whether we have a particular version of gene in ourselves can hint at, or present red flags to, a predisposition of various diseases like a cancer. When science first sequenced all the genes in the entire human genome, it became possible for scientists to start to compare the genomic patterns of larger groups of people, looking for more clues to health and disease. Finding out about the possibilities and consequences of getting a certain disease or condition can pose ethical and cultural quandaries for families. The job title ‘genetic counsellor’ became a thing in the late twentieth century and it's blossomed since then.
CLARA GAFF
Hi, I'm Clara Gaff, I'm an Honorary Professorial Fellow at the University of Melbourne and the Executive Director of the Melbourne Genomics Health Alliance.
CHRIS HATZIS
Professor Clara Gaff was awarded the Most Valuable Woman in Leadership in the Biomedical Space for 2019. She’s made a difference to people's lives by bringing genomics into healthcare. Professor Clara Gaff sat down with our reporter Dr Andi Horvath to talk about her work.
ANDI HORVATH
Clara, what's changed in your industry? I'm not sure how far we should go back because I know it's big, but let's start somewhere.
CLARA GAFF
Well, maybe if we start when I was training as a genetic counsellor, when it was really at the early stages of genetic testing still. I remember just early - towards the end of my training going to a meeting where the Huntington's gene had just been found. What caused Huntington's disease, a really serious neurological condition, had been identified. To go from that, I guess 25 or so years ago - I haven't actually counted too carefully - to now being able to offer, in healthcare, tests to look at thousands of genes all at one time I guess still stuns me. It's something I could never have imagined.
ANDI HORVATH
This has been the revolution of genetics for human genomes. We sequenced the whole entire genome and we started comparing it to others. What has surprised you most about the changes in the field?
CLARA GAFF
There have been so many changes. I guess in our work we focus a lot on the challenges and overcoming the challenges. Maybe one of the things that's surprised me is just how effective this technology is. We've always talked - and there continue to be limitations of any technology. As a genetic counsellor, when people are coming in for genetic testing we say, well we might find something, we might not. If we don't, that's uninformative. The chances really often were that we wouldn't find something, even though we expected that there was a genetic cause to the condition, and for many, many people we couldn’t offer a test. Now the rates of diagnosis using this technology surpasses what I had expected.
ANDI HORVATH
Let's take it from the patient perspective. How does it all work? I come to you, or a version of you that is a genetic counsellor, why have I come to you and what can you do?
CLARA GAFF
So in genomics, at the moment, it's very much about identifying the cause of a condition in a person with a medical condition. So we're looking at 16 different types of conditions where genomics may make a difference to the healthcare of that person. So as a genetic counsellor, mostly I've worked with people that are likely to have been born with a change in their genetic makeup that's caused their medical condition. So for that person coming in, they would normally see a doctor in genetics. That would be a medical geneticist, it might be experts in other medical disciplines. They would have an assessment. Like is this likely to be caused by something genetic? Is there a likely diagnosis or maybe many possible diagnoses?
Sometimes that one condition might be caused by many different genes. So that's one of the differences between genetics and genomics. With genetics, for example, cystic fibrosis, there is one gene that causes that and you look for mutations in that one gene. But there are many conditions that can look quite like each other. So there might be many genes that you might want to test. It was never feasible to test one gene after another. It was basically very expensive, very time consuming and not really feasible to offer within the health system. Whereas now patients can have one test to look at those many, many different possible genes that could be causing their condition.
So for the person, it's I guess maybe not terribly dissimilar to having a genetic test, in that a sample of blood is collected and that's tested. It takes quite a while in normal conditions, so three months or so to get the tests; this isn't straightforward testing at all. The implications that are discussed are similar still, because we're still talking about genetic conditions, inherited conditions. So whether it's a genetic test or a genomic test, those implications are the same. One of the differences we see is, some of the challenges in genetic testing are really scaled up with genomics because you're looking at more genes, there are more bases that could be changed, and a lot of those genes are new. We don't really understand fully the differences between normal - for lack of a better word - and an altered gene. It's a bit like being given a book to read in Russian, and you know it can make sense but you're not really fluent in Russian.
ANDI HORVATH
Give us some perspectives of the types of disorders? Because I really have no idea. Do I go to you if I've had a cancer diagnosed or if I've got a neurological condition? Or do I go to you as a parent who's hoping to have children but has concerns?
CLARA GAFF
So with the genomic testing, at the moment we're looking at most of those situations. So with children that have been born with serious problems, may be acutely unwell, for example, in intensive care, and they think that there's likely a genetic basis to that. So Zornitza Stark and Sue White, who are medical geneticists at the Victorian Clinical Genetics Services and Royal Children's Hospital, have been working with the intensive care unit there and the laboratories, to see if the testing, the genomic test, can be speeded up. So that instead of taking three months, it can be done much, much faster for those children to get a diagnosis more quickly and, therefore, to influence the medical care of those babies. That's been phenomenally successful. More than half of the babies that are tested are getting a diagnosis and I think three-quarters of those then are having a change in their care. Sometimes that's quite a simple change that can really make a rapid improvement.
ANDI HORVATH
Just to give me prospective conditions, these are neurological conditions like cystic fibrosis which is one gene, or Huntington's which is one gene? Some cancers are more than one gene?
CLARA GAFF
Yeah. So with conditions that are just one gene, that would stay with genetic testing. These would be conditions more - I'm not being specific about names because there are so many. They tend to be the ones that are seen more rarely. So each condition individually is very rare, but when you look together of all of these rare genes they can be more common.
ANDI HORVATH
I get it. So if you put together datasets of rare genetic disorders, all of a sudden they're not so rare because you've detected them.
CLARA GAFF
Exactly, yeah.
ANDI HORVATH
Okay, I'm starting to see the picture.
CLARA GAFF
So some of the changes - some of the patients that we focused on initially were ones where the clinician thought, well I think this is the likely diagnosis. Then the laboratory would sequence all the genes, but they would just analyse the genes that the clinician thought were the ones related to that condition. But if they didn't find a change causing a condition, then they could look more widely after that.
ANDI HORVATH
So by looking at that data, that's a domino effect. Because if there's someone else around the world who has something similar, then they can match it to a pattern they saw in Melbourne, to one in London and one in Beijing. So this starts to work as big data.
CLARA GAFF
That's exactly what it is and that's exactly the way it's working. So they might find a change in a gene that maybe one person in the world has been found to have. Then they can say, well, look we've got this patient here with a change in this gene as well. Then somebody else will say, oh we've just found one too. So suddenly around the world people are connecting and bringing together their knowledge to say, okay, well we think this really is then the gene that's truly causing this condition. That's actually the case for the families as well. Families are going on Facebook and they're forming groups for these conditions where there may only be five children around the world with it. They're forming groups and they're comparing their experiences raising their children, what their children are like, to work out, well, is this something caused by the condition or is this just my child.
So that's what we call germline conditions where people are born with a change, but this is also relevant for cancer. So some people develop cancer because they're born with a genetic change. Everybody's heard of the BRCA1, BRCA2 genes causing breast, ovarian cancer; Angelina Jolie made those famous. But for most people cancer happens because there's an accumulation of changes in cells that happen through their lifetime. Some of those changes can mean that the cancer is more likely to respond to particular medications.
So Jayesh Desai, at the Peter MacCallum Cancer Centre, and his colleagues are doing a project, through Melbourne Genomics, testing people that have got advanced cancer. So really where many of their therapies and treatments have been exhausted, to see if there are what we call actionable, or potentially actionable, changes. So an actionable change is where there may be a treatment on the basis of the changes that are seen in the cancer. So what they're doing is seeing if there are changes there that then people can go into cancer trials to test out the latest treatments, experimental treatments, to see if that will make a difference to their outcome.
ANDI HORVATH
Clara, you've been instrumental in bringing genomics to healthcare. I get it, there's a lot of big data out there, it just needs to be linked. How on...
CLARA GAFF
Yeah.
ANDI HORVATH
...earth did you herd all those cats?
CLARA GAFF
I was given very good advice by Doug Hilton who's the Director at the Walter and Eliza Hall Institute. Melbourne Genomics was put together by the CEOs of the Parkville-based institutions initially; we now go more broadly than that. His advice was do a vision; what would genomics look like in five years' time if people were working together, what would it look like? That really helped, as you say, herd the cats. Because it meant that we weren't saying to people, well you have to change what you're doing now. What they were doing at that time was sequencing lots of genes, but not all genes or the entire genome. They might sequence, for example, the 50 genes known to cause inherited cardiac conditions, for example. So we were saying, that's fine, we're not asking you to change anything. What we're doing is testing out a vision for the future for five years' time. We're testing that vision out and are you willing to try that out? They said, yes, we'll do that.
So as well as doing their regular work, everybody worked very hard, and I think was very generous with their time and enthusiasm, to test out an approach where we were providing genomic testing to patients - and when I say we, it's the clinicians, the laboratories, everybody working together - as part of medical care. But unlike everywhere else in the world where people were testing it out after every other test had been done, we were saying, well, maybe in healthcare this would be the first test that you will do. So we will test it out as the first test, but in parallel, so at the same time as people are getting their standard care. So their usual care is not being compromised or changed. What we're doing is adding an extra test on early. Then what we can do is compare the two; what happens with the usual care and what happens with the genomic test. So we could do a direct head-to-head comparison for that individual and then across that patient group as a whole, we'll say this is what a difference it can make to this group of patients.
So we funded the test and the laboratories introduced this technology. I think this really helped the laboratories accelerate getting this technology into the labs sooner, rather than keeping on just doing panels for longer. Then what we did was not direct what should happen, like the medical care obviously of the patient, that was for the clinician to decide. But then we looked at what decisions the clinicians were making on the basis of this information. So did it change patient care or not and then what were the outcomes out of that. Very importantly - you know I talk about doctors and laboratories, but really at the heart of this are the people with the medical conditions. We also surveyed them, interviewed them, and we have a fabulous community advisory group who make sure that patients remain at the centre.
This is another difference when we started at the time, people were very focused on the technology challenges. So they were I'd say technology driven, lab driven, this is all the stuff that the technology can do, look at this, isn't this great, and focus on those challenges. Where we turned it around the other way and said, actually, being a genetic counsellor and seeing the challenges of getting new tests into clinical practice, the challenge is actually going to be at the healthcare end. Having a new technology available doesn’t mean that it's necessarily taken up in healthcare by clinicians and patients. So our focus was on the healthcare part and with clinicians and patients and that interaction. Saying not so much that genomics is going to - people talk about it being paradigm changing. It's not going to be paradigm changing when you're talking about that individual clinician-patient interaction. It's another investigation and there are principles of good clinical practice and let's apply those.
So an example of that is when one of the ethical challenges and barriers people were saying was, well we can sequence all of these genes, maybe we should be looking at absolutely everything. So that would mean that somebody coming in with a neurological condition, if all of their genes were examined, they could potentially find out that they had a variant or a mutation causing breast cancer, but that's not what they were there for. So we turned it around and said, well when a patient comes in, you don't do a whole-body MRI, you do an MRI of the part of the body that you want to look at. So it's the same with genomics, or it should be the same with genomics. Let's test out an approach where, yes, everything - the data is there, all the data gets generated, but you're only analysing the genes that are relevant to the person's condition.
Clinical genes, not highly speculative research genes that - and this comes back to the big data and the sharing as well. That data we will make available, with the patient's consent, to researchers to then look for other genes on a research basis, but we're focused on clinical care.
ANDI HORVATH
There had been a lot of discussion in particularly the early twenty-first century about privacy in genes, about insurance companies making decisions on your predispositions and it was seen as an invasion of privacy. There's less talk about that there. Is that because of privacy legislation and protection of individual's data has come into place here? Or what's happened there?
CLARA GAFF
There's sort of two slightly different situations. So with the people that are being seen through our program, they have medical conditions. So with insurance - and we're fortunate in Australia health insurance is community rated. That means that an individual's personal health history isn't taken into account in the premiums. So when we're talking about concerns about privacy and health, it applies much more to life insurance, income insurance, trauma, those forms of insurance. So the people that we're seeing already have a medical condition which they need to declare. What this testing is doing is looking to understand the cause of that condition. So that shouldn't have an impact on that person's insurance.
The concerns lie more with people who are well, who are having testing to look for their future risk of disease. Certainly people that are having what we call predictive testing - and largely this has been, not only, but say hereditary cancer, hereditary cardiac conditions - that people are well and they're learning they've got a higher risk of developing cancer. So they are required to declare that to their life insurance company or, yeah, the company that they're getting insurance products from. So the concern has been that companies will base their - will deny people that are found to have a mutation that's highly likely to cause a condition in the future, that they won't be able to get insurance.
ANDI HORVATH
What is the cost-benefit analysis from doing genomics? You talked about an earlier pilot study which was in colorectal cancer, wasn't it? You said, right, I want to look at how it looks five years from now, how genomics can help in healthcare, does it help in healthcare? It did help in healthcare. So what was the cost-benefit analysis that you examined?
CLARA GAFF
Colorectal cancer was one of the ones we looked at early on. Across our whole program there have been 16 different conditions that we've looked at. Some of those we haven't found any improvement over existing testing. In fact, colorectal cancer was one of those. So in hereditary colorectal cancer there was a very good panel test. So they would just sequence the genes known to cause hereditary cancer, and the genomic test was not an improvement over that. So that's really helpful. That tells the hospitals and funding bodies, government, well this is okay as it is.
In the other example I gave, what we call childhood syndromes, so children born with physical features that suggest that they have an inherited condition, what we did was compare the costs of their usual investigations with the costs of the exome or the genomic test and we compared the number of diagnoses that were made. We found that there was, at that time, five times the number of diagnoses with the genomic test at one-quarter of the investigation costs. In fact, for each additional diagnosis that was made over standard care there was actually a cost saving. So that's very powerful figures.
That was part of the submission to the Medical Services Advisory Committee for that testing to be included on the MBS, item number, and the recommendation back to government has been that that should be now a funded test. I mean that's just really looking at that this is a better investigation, like more effective at making a diagnosis. We also try and really understand the benefits that come to people individually that might not be captured by the health technology assessment. The kind of certainty that having a diagnosis makes. Even having a diagnosis can make things easier with the NDIS in terms of application to the NDIS, the certainty parents then have around what the cause of their condition is, I mentioned earlier linking up with other families around the world that have children with the same condition.
ANDI HORVATH
So not just the medical healthcare, but also social, cultural support for conditions, and creating communities which would have never have found each other otherwise. I really like this idea of almost customised healthcare that actually depends on everybody playing almost on a global scale.
CLARA GAFF
Yeah.
ANDI HORVATH
That's kind of the nice side of this story.
CLARA GAFF
It's been something that's been really evident in genomics. So we've seen it here locally with researchers, clinicians, laboratories working really closely together. The challenges of this big data are such it really needs people working closely and bringing the different expertises together. We see that around Australia. I've talked about Melbourne Genomics. There's Australian Genomics Health Alliance which is bringing the state-based activities and working on national harmonisation. Then linking in internationally with groups like the Global Alliance for Genomics and Health. So they're all about promoting sharing of data and sharing experience around the world. So there's very active collaboration. Everybody understands that nobody can do this alone. Nobody's going to be able to understand enough of the genome to make enough of a difference.
ANDI HORVATH
You deal with many, many stakeholders, from families, to genetic conditions, to communities of rare diseases, to scientists, researchers, large organisations, corporations. That's quite an extraordinary challenge. What drove you to make those connections? Most people would look at that and go, oh too hard.
CLARA GAFF
Yeah, that's a very good question, what drove me. I don't know that it's so much something - what drove me? I guess I've always worked across organisations. Coming out of a laboratory background and then going into healthcare, I right from the start in genetic counselling saw how necessary it was to work closely with the laboratories to understand the test results that were coming out, and equally for them sometimes to understand some of the challenges. That the way they might want things ideally to work into the sample collection might not always work for other reasons. Then sometimes with research colleagues, some not understanding why research findings aren't immediately being taken up in healthcare. Because they've done this fantastic research; why aren't clinicians just embracing it?
This is one of the big challenges. Often people talk about technology and the challenge of the technology and the challenge of the big data. But, actually, healthcare is designed to be safe and secure and reproducible as much as possible, whereas science is all about innovation and cutting edge. We often use the analogy of the fighter plane. Science is the fighter plane; innovative, pushing further faster. Whereas healthcare is more like getting on a plane to go to Sydney; you want it to be safe, you want to know when you're going to arrive, you want to know how much it costs, you want good value, you want it to be the same each time. So healthcare is not designed to change very rapidly on a whim. It needs careful thought.
In fact, most - I think the RAND Corporation found that to get a new technology into healthcare takes about 15 years. We've really accelerated that, because this is now five years or so. That's because we've taken a very broad approach, broad in terms of types of conditions, broad in terms of types of organisations, broad in terms of the strategies that we're using. So we talked about flagships, those 16 conditions, we talked about the flagship projects. They're working on getting evidence to put back to hospitals and to government of when this testing is worth doing and how it should be done. The how is not just in the laboratory, but the how is also the talking with people about it, the what information needs to go to the laboratory, how it can be discussed with patients, who in the clinical services should be doing this.
The clinicians that are running these flagships, they're getting the hands-on experience of doing this and evaluating it. So they're then the leaders within their discipline that can help their colleagues not be over-enthusiastic and, equally, not be too pessimistic, but to get it just right. So there's that level of change. There's change that needs to happen in terms of clinicians using tests, the hospital services being able to introduce this safely within their organisations, and then at the government level with the policy decisions and funding decisions about when and where and how testing should be done.
ANDI HORVATH
Go, Clara. You're making a difference.
CLARA GAFF
Lots of people are making a difference. I mean we think we've served as a - my team, I have a fabulous team. My team work really hard with all of those stakeholders that you mentioned. Really, it's about what we do is really catalyse that collaboration of people working together and help that keep on happening, and provide the opportunity, through the funding that we have from the Victorian State Government and the member organisations, that provides the opportunity for people to work together, to learn together and to make these changes. Without that, people would be in silos, the same lessons would be learned over again in different hospitals. What happens at the Austin Hospital and the Monash, they wouldn't be learning from each other necessarily. So this is really helping everybody learn, share experience and ensure that wherever a person goes, whichever the hospital they go to, they're getting access to good genomic care.
ANDI HORVATH
Clara, do you still deal with some misconceptions from various groups about what's possible with human genomics and healthcare?
CLARA GAFF
So I think there's been a lot of hype about genomics. So people would get very excited about all the possibilities. We see that sometimes, when testing is first made available, that this is going to be the answer for everything. But genomics is just part of the puzzle; it can give some answers, it's not going to give all answers. You talked about data. Again, genomics is the start. There's proteomics, epigenomics. I think over time we'll understand more and more of all of these different parts and the information will be combined together.
ANDI HORVATH
Of course, genomics being just the gene, proteomics being the proteins that flip around. What was the other one?
CLARA GAFF
Epigenomics.
ANDI HORVATH
Which is the environment in which the genes and the proteins operate.
CLARA GAFF
Yeah. So one of the things I think that sometimes people can have a misconception about is that we'll be able to give much more definitive answers or much more - be able to give…
ANDI HORVATH
Detailed, is it?
CLARA GAFF
Yeah, more detailed or to know with more certainty. Because we're looking at - inverted commas - all the genes, that therefore we can provide more information about what the course of the disease will be. Now, that might be the case for some changes, but not others. I think people expect that we know a lot more about the genome than we actually do. There's still a lot of uncertainty there. This is why the sharing around the world is so important. To really develop a much better understanding so that we can give - there are lots of what we call variants of uncertain significance. So some changes we can say, well that's clearly not causing the condition, the laboratory can say that. But others we can say, well, yes, we've seen this many times and it biologically fits and it meets the criteria that this is likely to be disease causing. But then there's a lot of variants that fall in the middle that are suspicious, but there's just not enough known about them yet to be certain that they are definitely disease causing or definitely not disease causing.
ANDI HORVATH
It's not a yes or no, it's a grey area.
CLARA GAFF
Exactly. There's a lot of grey still.
ANDI HORVATH
All right. No doubt you've seen the film, Gattaca. For those of you who haven't, you might want to check it out. What's your response to the comment, we are our genes?
CLARA GAFF
No, I'm not a - [someone who would say] it's all about the genes. I think our genetic makeup is an important facet, but there's so many more other aspects, like environment, that make a difference.
ANDI HORVATH
We can rise above our genes.
CLARA GAFF
Yeah. Genetic information provides predictive information, sometimes in terms of what might happen in the future, like what diseases might happen. But it's predictions and not certain information.
ANDI HORVATH
Right, so we're not our genes. Our genes are our predictions.
CLARA GAFF
Yeah. Our genes are maybe hints or glimpses, but certainly not the full picture.
ANDI HORVATH
Clara, I bet you a lot of people ask you about those ancestry.com ads you see on television. They seem to be a thing at the moment.
CLARA GAFF
Yeah, ancestry testing is very big and I think for many people it's a fun thing to do. I cringe slightly when they link in their interests with their genetic makeup and what their long-distant ancestors did. I think people look to make meaning of things and that's totally fine. But again as a genetic counsellor, I think those aren't necessarily in your genetic makeup.
ANDI HORVATH
They're not scientific.
CLARA GAFF
No.
ANDI HORVATH
Your Viking that rode the seas doesn't mean that's why you like surfing.
CLARA GAFF
Sylvia Metcalfe has been doing some research I'm part of, with people doing what we call online testing. People are doing ancestry testing and then they're having it reanalysed for so-called health purposes. But the evidence base behind those associations and the ability of that testing to predict future health is very variable. I think while people see it as a nice added extra, in fact it's something to be careful of and to really look carefully into.
ANDI HORVATH
Yeah. Hey, we all evolved from Eve in Africa, so just stick with that. We're not sure her hobbies were the same as ours.
CLARA GAFF
Yeah [laughs].
ANDI HORVATH
You can't make that interpretation. Next time we hear the word DNA being bandied around to describe a corporation or a car, what would you like us to think about?
CLARA GAFF
I think it's good that people have a sense that DNA is something that's intrinsic to us. I hope that people don't take that in a very literal way. It's in my DNA is something that's often said by people to mean it's something really fundamental and important and intrinsic to me. But as a genetic counsellor, I think but it's not literally caused by your genetic makeup. So I think it's that difference of something intrinsic, but actually things that get talked about as in my DNA aren't necessarily actually genetic related.
ANDI HORVATH
Sure. I mean in some ways DNA only works if it's in a certain environment. I love that complexity. Professor Clara Gaff, thank you.
CLARA GAFF
Thank you, Andi, it's been an absolute pleasure.
CHRIS HATZIS
Thank you to Clara Gaff, Honorary Professorial Fellow at the University of Melbourne and the Executive Director of the Melbourne Genomics Health Alliance. And thanks to our reporter Dr Andi Horvath.
Eavesdrop on Experts - stories of inspiration and insights - was made possible by the University of Melbourne. This episode was recorded on May 16, 2019. You’ll find a full transcript on the Pursuit website. Audio engineering by me, Chris Hatzis. Co-production - Silvi Vann-Wall and Dr Andi Horvath. Eavesdrop on Experts is licensed under Creative Commons, Copyright 2019, The University of Melbourne. If you enjoyed this episode, drop us a review on iTunes and check out the rest of the Eavesdrop episodes in our archive. I’m Chris Hatzis, producer and editor. Join us again next time for another Eavesdrop on Experts.
Genomics is a rapidly evolving technology that can help identify the genetic cause of a condition in a person. It can also find a person’s predisposition to various diseases like some cancers.
When science first sequenced all the genes in the entire human genome, it became possible for scientists to compare the genomic patterns of larger groups of people – looking for more clues to health and disease in the ‘big data’.
Professor Clara Gaff was awarded the Most Valuable Woman in Leadership in the Biomedical Space for 2019 for her work as Executive Director of the Melbourne Genomics Health Alliance.
And the way genomic technology has advanced still stuns her.
“The rates of diagnosis using this technology surpasses what I had expected,” she says. “It’s something I could never have imagined.”
Episode recorded: May 16, 2019.
Interviewer: Dr Andi Horvath.
Producer, audio engineer and editor: Chris Hatzis.
Co-production: Silvi Vann-Wall and Dr Andi Horvath.
Banner image: Shutterstock
Subscribe to Eavesdrop on Experts through iTunes, SoundCloud or RSS.