Ep20: Genomics for the Generalist – Part 1
Ep20: Genomics for the Generalist – Part 1
To play the media you will need to either update your browser to a recent version or update your Flash plugin
This is a two-part series looking at how modern genomics is changing clinical practice, and how a physician can hope to keep up with the pace of discovery and technological development. Some single gene tests and gene panels have been available off-the-shelf for years, but whole genome sequencing is becoming more accessible and affordable every day.
In the first episode we discuss the differences between these technologies in terms of cost and practical utility, using diagnosis of Mendelian conditions and rare developmental conditions as examples. We also talk about counselling parents through prenatal or preconception screening, and the psychological burden of genetic diagnoses. The potential of precision medicine and pharmacogenomics is also covered. Our second episode, published in tandem, begins with the question of disease risk and how to present uncertain predictive diagnoses.
Guests: Associate Professor Kristine Barlowe-Stewart FHGSA(GenCounsel) (Director, Master of Genetic Counselling Program, University of Sydney), Professor Leslie Burnett FRCPA, FHGSA, FCAP (Chief Medical Officer, Genome One), Dr Michael Gabbett FRACP (Royal Brisbane and Women’s Hospital), Dr David Thomas FRACP (Director, Kinghorn Cancer Centre; Director, Cancer Division, Garvan Institute for Medical Research).
Links to resources relating to this two-part series are provided below. Fellows of the RACP can claim CPD credits for listening and further reading on this topic via MyCPD.
Genetic Services in Australasia
Clinical Genetic Services in Australia by State [Centre for Genetics Education]
Genetic Health Services in New Zealand [Genetic Health Service]
Genetics Fact Sheets [Centre for Genetics Education]
The NSW Genetic Counselling Workforce [The Sax Institute]
Genome One [Garvan Institute of Medical Research]
ACMG Recommendations for Reporting of Incidental Findings in Clinical Exome and Genome Sequencing [American College of Medical Genetics and Genomics]
Recommendations for Returning Genomic Incidental Findings? We Need to Talk! [Genetic Medicine]
Incidental Findings in Clinical Genomics: A Clarification [Nature]
Psychological Impact of Genetic Testing for Huntington's Disease: An Update of the Literature [Journal of Neurology, Neurosurgery, and Psychiatry]
Psychological Impact of Genetic Testing for Cancer Susceptibility: An Update of the Literature [Psycho-Oncology]
Other Journal Articles
Association of Biomarker-Based Treatment Strategies with Response Rates and Progression-Free Survival in Refractory Malignant Neoplasms: A Meta-Analysis [JAMA Oncology]
This episode was produced by Mic Cavazzini. Music from Blue Dot Sessions (“Cloud Line”), Chris Zabriskie (“Is That You or Are You You?”), Alex Fitch (“Celeste”), Cory Gray (“Terminal Two”), and Kromatic (“Club Crunk for Monkeys”); photo courtesy iStock. Pomegranate’s executive producer is Anne Fredrickson.
Editorial feedback was provided by RACP Fellows Dr Pavan Chandrala, Dr Tessa Davis, Dr Rebecca Grainger, Dr Michael Herd, Dr Paul Jauncey, Dr Joseph Lee, Dr Marion Leighton, Dr Anutosh Shee and Dr Ellen Taylor, and Advanced Trainee Dr Katrina Gibson.
DAVID THOMAS: I don’t think it’s quite like every other area of medicine, and we’re going to have to evolve at a legislative and ethical and social level and have more debate about how we, as a society, want to deal with that information. It’s so profound that I don’t think there’s an area of society that it doesn’t in some way touch.
MIC CAVAZZINI: Welcome to Pomegranate, podcast of the Royal Australasian College of Physicians. I’m Mic Cavazzini, and this is the first of two episodes looking at the place of genomics in clinical practice, and how a physician can keep up with the pace of discovery and technological development.
Sequencing the first human genome took 23 labs and 13 years, at a total cost of about three billion U.S. dollars. Now a single laboratory can churn through a person's DNA sequence in two weeks, and new disease markers are being identified all the time. Some people think this heralds a new era in 'precision medicine'—treatment tailored to an individual's unique genotype. Others are concerned the technology will uncover genetic risk factors that patients weren’t looking for and didn’t want to know about.
We’ll deal with that issue in the next episode, as well as the application of cutting-edge research into practice in the field of cancer genetics. But today we’ll start with some first principles with Professor Leslie Burnett, Chief Medical Officer of the Genome One clinical sequencing facility at the Garvan Institute of Medical Research:
LESLIE BURNETT: 20,000 genes in the human genome: we don’t know the function of all of those. So far we have a Mendelian base of about a third of them. Round about the 6,000 out of 20,000, we have a phenotype for which molecular basis is known, and probably we’ve found the relationship between the gene and the condition in half of those.
MIC CAVAZZINI: Mendelian conditions are those phenotypes that can be linked to a change in a single-gene, and whose inheritance can be traced through a family pedigree. One familiar example is cystic fibrosis, a disease that affects the lung mucosa. Those patients who have access to daily medical and physical therapy often survive childhood, but to a predicted life span of only 37 years.
The CFTR gene underlying cystic fibrosis was cloned in 1988, and targeted testing for mutations soon followed. There are hundreds of known mutations in this gene, the most common of which is so drastic that it stops the protein from ever folding into its functional shape. But clinical geneticist Michael Gabbett explains how outcomes can be drastically improved in individuals who carry a more subtle variant.
MICHAEL GABBETT: I’m Michael Gabbett. I’m a clinical geneticist at the Royal Brisbane and Women’s Hospital in Queensland.
So there’s a number of mutations, over a thousand in the CFT R gene, which lead to dysfunction and cause cystic fibrosis. Now, one of the more common mutations is one called G551D. This means that an individual produces the CFTR protein, just the protein doesn’t work quite as well as it does normally, and there’s medication which is now available to help that protein work better. So, if you find that someone has cystic fibrosis due to a G551D then that certainly changes management. It’s a bit of a success story in recent years.
MIC CAVAZZINI: Cystic fibrosis is a recessive condition, meaning that a mutant variant or allele must be inherited from both parents for an individual to express the phenotype. In people that carry only one mutant allele, the normal copy generates enough functional protein to mask the effects of the mutation.
However, some genes are so critical to a molecular pathway that a mutation in even one allele of the pair is pathogenic. These are known as dominant conditions—Huntington's disease being the classic example. Testing for many Mendelian conditions has been going on for decades, but more and more conditions have been described which cannot be resolved by a single-gene analysis.
LESLIE BURNETT: So, if you know what you’re looking for, you know a condition is caused by one gene—you look at the one gene. But increasingly, conditions are known to have a number of genes that may influence them. It may be a pathway where different parts of the pathway can be affected, it may be a condition where a number of contributing genes. So one bundles different groups of genes together into commonly-requested combinations and that’s basically a panel. You can buy off-the-shelf now a number of commercial kits that will do the most frequently requested genetic tests and simple panels. If you find your answer, great.
The second approach would be that if you drew a blank on the narrow panel you might look at an extended panel; these are genes that there’s a looser association or a looser correlation, the causation is less clear. And then you’ve got the third approach which is to look at the whole genome, and it may be more convenient for the patient, simpler for the diagnosis, faster, cheaper, to look at the collection of genes and look at it in one go.
MIC CAVAZZINI: The technique of exome sequencing involves pulling out only those regions that code for protein, which makes up one percent of the entire genome. But 70 per cent of the genome is actually transcribed to RNA, and it appears much of this is involved in regulating the timing and location of gene expression.
Therefore, whole genome sequencing is used to process almost every strip of DNA—but you still need to choose which of the 20,000 genes to focus on in the bioinformatic analysis. The advantage of collecting all the data in one go is that you’re then able to reassess different panels of genes with each new diagnostic hypothesis. It's also much easier to adjust an analytic panel to the latest research findings than it is to develop a new sequencing array and have it validated for clinical use.
LESLIE BURNETT: So, we are starting by focusing on assisting with the diagnosis of rare or difficult-to-diagnose conditions that are suspected to be genetic but have, because of lack of knowledge, not yielded to the application of traditional genetic testing. Sometimes you may have 10 or 20 or 50 or 100 genes that may be involved. You can do it all in one go. Or you may want to do it once, and reinterrogate over time. So the beauty about the whole genome approach is you can actually apply, almost hypothesis-free, you can find diagnoses that weren’t suspected or weren’t even discovered at that point and yet being able to produce a diagnosis.
One of our early cases that was referred to us was the case of a young child who appeared to have some form of immune disorder, such that their platelet levels were falling and the child was extremely ill. And so we sequenced the child and other family members and we were able to find that indeed there was a fault in a plausible gene that might explain this condition. But what was interesting is that this gene lay on a pathway for which there was a medication available—but not for this particular presentation, but rather a quite different medical condition. So the clinicians decided that it was ethically defensible to apply for permission to apply this medication and the child left ICU to watch the opening of Star Wars only some two weeks later. And I think that’s one of the remarkable success stories and our dream, it was almost our very first patient coming through.
MICHAEL GABBETT: Paediatrics is probably where a lot of the clinical genetics grew up in. It’s not all about genetic testing; so there is clinical acumen involved as well. So, we do have children that walk through the door and they have a pattern of facial features that you recognise, and if you’ve got a very well-defined clinical phenotype and there are a number of genes associated with that phenotype then that might be all you look at. Certainly if there is no obvious condition apparent then there are, I guess, what you could consider screening tests. The literature says we’re probably getting a diagnosis, say, 25 per cent of the time if they come through the door cold. But the things I see are rare and esoteric; we’re talking about conditions where there are only a handful reported in the world. Having said that, I’ve certainly witnessed conditions where they’ve previously been unreported and a colleague has described it and then all of a sudden they sort of fall out of the sky and everyone else starts to recognise it.
MIC CAVAZZINI: Some developmental conditions result from incorrect replication of chromosomes in the gametes, and are easy to detect by techniques such as karyotyping or fluorescence in situ hybridisation. Examples include Turner syndrome and the microdeletion syndromes. But even where diagnosis of a genetic condition doesn’t lead to treatment for a given paediatric patient, the benefit to the family can be tremendous, says Associate Professor Kristine Barlowe-Stewart, Director of the Master of Genetic Counselling program at the University of Sydney.
KRISTINE BARLOWE-STEWART: I’m sure you’ve heard about genomics ending the diagnostic odyssey. What parents want is an answer. They need to know, “Did I do this to my child? Was it my fault? Was it something that I did in pregnancy?” And so, having a genetic diagnosis relieves that guilt, that blame that was often there. So it’s important for a whole lot of reasons, it can affect how they feel about themselves and their family.
MICHAEL GABBETT: The prenatal question, that’s one of the big questions we’re asked, what’s the chance of this happening again if this couple were to have another child? I guess from a medical point of view it’s important to give people answers, give them a sense of owning their own destiny in a way, empowering people to plan for their future. Now for some couples that’s not an issue, they’re quite happy, and if fate affords them another child with a developmental disability they’re not that fussed on necessarily getting an answer.
KRISTINE BARLOWE-STEWART: The ideal, of course, is preconception screening—what we’ve done in the Jewish community with the high school program; I’ve been involved in that since 1995. You know, there has been no child born with Tay-Sachs Disease in Australia to any couple screened.
And so, you know, we would really encourage taking a greater, more in-depth family history and updating it as well. But many of these recessive conditions like cystic fibrosis there is no family history, so asking about ancestry is very important. Like with Tay-Sachs, like the whole range of other conditions that are common in the Jewish community. If you’re of Caucasian background then testing for cystic fibrosis would be appropriate. If you are of Mediterranean, Southeast Asian, Chinese, then thalassemia screening is very important as well.
MIC CAVAZZINI: The entire genome is about 3 billion nucleotides bases long, and any two individuals differ from each other at one position in a thousand. To study the sequence genome the lab starts by breaking the chromosomes down into fragments of a few hundred kilobases that can be read with consistency by the biochemical reagents. The computational challenge is to put the snippets back together like a colossal jigsaw puzzle. There are many stages in the process where errors can be introduced, and the Genome One facility is only the second in the world that has been accredited to meet the national standards of a clinical pathology service.
It’s an impossible task for a single person to be on top of all the latest advances in the technology and the genetic research. Oncologist David Thomas explains what the general physician needs to consider during a consultation that might require genetic testing.
DAVID THOMAS: So, my name is David Thomas. I’m Director of the Kinghorn Cancer Centre, I’m Head of the Cancer Division of the Garvan. I’m a medical oncologist with a specific interest in the applications of genomics to cancer care.
So, I imagine that the average physician needs to be able to separate out the conventional from the unconventional and then where to go when it gets unconventional. And that’s about referral patterns. And the vast majority of things will be canonical and, you know, there’ll be Stage 1 breast cancer which needs a surgeon and it needs chemotherapy to reduce the chance of the tumours coming back, and that will be perfectly standard.
It’s when the woman comes back the third time with advanced disease and you need to understand, now that the chemotherapy has stopped working, which test do you perform, what does it mean when you get the results and what do you do next? It’s about knowing who to go to who may have a specialist interest and provide advice. And I mean that’s part and parcel of clinical practice, I guess.
MICHAEL GABBETT: The generalist is I guess the phenotyper, so the most important job is: “Look, I think this is a genetic condition, these are the features. There might be a differential diagnosis; could you do a genetic test?” And that’s where, increasingly, genetic pathologists and clinical geneticists can help send the DNA off to the right lab and know what’s available locally or internationally, what’s the cheapest test, what the implications of a test may mean for other family members.
Genetics obviously involves your whole body; you can’t hold that knowledge in your head on every single genetic disorder. A lot of what we do is work with organ-specific specialists. Now, you’re not expecting your average specialist to be reading Nature Genetics, of course, but specialists will learn about these discoveries in the following year or two and they bring this knowledge to the genetics clinic. You often find that in the larger units that clinical geneticists do take up special interests. Two of my colleagues, one’s interested in cardiac genetics and one’s interested in renal genetics, and it’s not uncommon for us to get on the email and email the world expert on X, Y or Z for their opinion.
KRISTINE BARLOWE-STEWART: But people still see their doctor as the one to go, they want to talk to someone. And so we have to make sure that the doctors that they go to see can answer the questions they will increasingly ask. And the genetic counsellors are going to be pivotal in liaising with non-genetic specialist health professionals.
So, I founded the Centre for Genetics Education and on the website is every genetic service in Australia and New Zealand. The Human Genetics Society of Australasia also. Some states in Australia have a sort of a hub model and they would fly out to do clinics. And there are three main hubs in New Zealand, two in the South Island and one in the North. New South Wales has always been a bit different, in that we have genetic services in a lot of the major hospitals, but there is a huge workforce shortage.
MIC CAVAZZINI: Another outcome of next generation genomics is precision medicine—the idea that drug or lifestyle interventions could be tailored to a person’s particular gene profile. For example, blood-pressure lowering medications like ACE inhibitors are less effective in people that carry mutations in the renin-angiotensin-aldosterone pathway. And allopurinol therapy for gout can actually cause adverse cutaneous reactions in those that carry a variant of the HLA-B gene, common to East Asian populations.
David Thomas explains how pharmacogenomics could be used to find more appropriate medications.
DAVID THOMAS: So there’s certainly lots of examples where a drug’s response or efficacy will be determined by a genetic background and differ between populations. Another example of a drug reaction is the gene dihydropyrimidine dehydrogenase, which is important in metabolism of the drug 5-fluorouracil—which is a 40, 50-year-old chemotherapy used frequently in, for example, bowel cancer. And people who have got mutations in that gene have really dire, life-threatening side-effects to a drug which is generally innocuous. So you could imagine that in some future if we all have whole genomes at birth, that the clinician who had that information for every patient on their computer would say, “I need to treat somebody with hypertension.” You would then get an alert that comes up saying, “This person has a variant and the renin-angiotensin pathway which makes them resistant to this particular drug. Use drug X instead.”
MIC CAVAZZINI: Genetic Pathologist Leslie Burnett, once more:
LESLIE BURNETT: And certainly we will be very keen to explore the move from curative or treatment to preventive medicine—well, it’s already happening now that a number of people are having the genome sequence at some point early in life, and we would imagine that somewhere between one and five per cent of the population will be carrying some variant that is going to potentially put them at increased risk of a condition they might otherwise not have been aware of, and so those patients will be able—well, not patients yet—those referrers potentially could become patients at that point and might be able to be offered early monitoring or early intervention so they never will actually get those conditions. And that is, I think, one of the enormous potentials yet to be explored. Then having the extraordinary tool of having one’s genome and being able to re-interrogate it, it really opens one’s eyes to the possibilities of what may be done down the track. Some of this is hype and some of this is prediction and some is reality, and I’m sure the future is not as far in the future as many of us would think it might be.
KRISTINE BARLOWE-STEWART: Of course, if you find a hereditary cancer then who looks after the unaffected relatives? That is a big issue in terms of health delivery. Physicians, GPs, other specialists, they’ve got sick people in their waiting rooms. You’ve got a well woman coming in. Will they want to see them? That’s the ripple effect in terms of health delivery.
MIC CAVAZZINI: Genetic counsellor Kristine Barlowe-Stewart describes how to manage the anxiety generated by handing out genetic diagnoses to otherwise healthy people.
KRISTINE BARLOWE-STEWART: People certainly are affected psychologically in the short term, but that does not last—for the vast majority, 12 months later people have incorporated that information into their lives. And as part of genetic counselling we make sure that people are as well prepared psychologically for this information as they can be. And we then support them and we’ll refer them for extra assistance. But the bottom line is the pre-test preparation is so important. You’ve got to manage expectations, prepare them for uncertainty, prepare them for unexpected findings, prepare them for no answer.
MIC CAVAZZINI: And does that message, that pre-test message that you give, change very much depending on the condition they’re interested in and whether it’s treatable or not?
KRISTINE BARLOWE-STEWART: Yes. If it’s treatable the interest in it is higher, but that doesn’t mean to say that people still won’t want to know even if it’s not treatable, because people like to prepare their lives. When the test for Huntington’s disease, when we first isolated that gene and it was possible to test for it, 60 per cent of people on average said, “Yes, I want it.” Only about 20 per cent of people turned up to have it. But some people turned up years later, after they’d thought about. So you cannot look at a person and say they will or they won’t.
MIC CAVAZZINI: And as a counsellor you’re agnostic about what the best strategy for wellbeing and psychological preparation is. You can’t say what is the best approach?
KRISTINE BARLOWE-STEWART: No, you can’t. It’s called patient-centred counselling. You just work with a patient in front of you, walk alongside them.
MIC CAVAZZINI: Whole genome sequencing is not directly funded on either side of the Tasman. Australia’s MBS only lists specific tests for haemochromatosis, Factor V Leiden thrombophilia, protein C or S deficiencies, and antithrombin 3 deficiency. As well as diagnostic testing of a patient, screening of a first-degree relative is also permitted. In New Zealand, the Genetic Health Service is federally funded, and will assess referrals for a much broader range of conditions.
Michael Gabbett describes the comparative costs of analysing gene panels from a whole genome sequence.
MICHAEL GABBETT: I guess it’s a bit of a “you get what you pay for.” So 4,000 dollars is probably what the diagnostic labs in Australia would be charging to look at a single person’s genome and that would be the whole genome. I certainly can have a next-generation sequencing and have a panel done for, say, around 1,000 dollars. So I don’t need all that bioinformatics done to look at the rest of the genome, I just want to look at neuromuscular genes, for instance, and that will cost a lot less.
So, the bioinformatics is what costs a lot because that’s bringing in an actual person with the expertise. As good as computers are you still need a human brain to sort of look at the data and say, “Oh look, you know, our whole genome sequencing did not find a mutation in this gene, but I know from experience that this exon actually isn’t read very well from next-generation sequencing. We might have to go back and look at more traditional ways of gene sequencing just for this exon.”
LESLIE BURNETT: It’s not a mental image of pouring DNA in one end of a sequencer, pressing the button and out comes your diagnosis at the other. These are massive computational problems, and for very complex conditions, particularly if the gene has only recently been described and you’re not sure whether the variant is causative or not, you may need to look at how it associates with affected or unaffected individuals within the family. So, clearly the cost will go up because you’re looking at more and more individuals.
At the moment, the two sources of funding are the hospital or the health entity, as is the case of many other genetic tests that aren’t on the Medical Benefits Schedule. And in some cases the patient is not eligible as a public patient, and so the patient of a family is electing to fund it themselves and pays for the cost of the testing. But there are a number of conditions for which we’re trying to draw attention to the government that you can only make these diagnoses by whole genome sequencing—and a mechanism needs to be found to fund this.
MIC CAVAZZINI: That was part one this Pomegranate series on “Genomics for the Generalist.” Thanks to Kristine Barlowe-Stewart, Leslie Burnett, Michael Gabbett and David Thomas for their contributions. The views expressed are their own, and may not represent those of the Royal Australasian College of Physicians.
For resources mentioned in the podcast, or to claim CPD credits for listening, visit the Pomegranate website at racp.edu.au/pomcast. You’ll also find our contact details if you have any feedback, and please share the story around using the hashtag #RACPpod.
I’m Mic Cavazzini. Please download the next episode to hear the rest of the story.