Genetic sequencing to diagnose rare diseases

Rare diseases are surprisingly common - about 30 million EU citizens are currently living with them. As each condition typically only affects a handful of people, reaching a conclusive diagnosis can be akin to finding a needle in a haystack. Sarah Williams discusses the challenges and opportunities of genetic sequencing for diagnosing such conditions with Lucy Jenkins, interim director of the Regional Genetics Laboratories and consultant clinical scientist at the North East Thames Regional Genetics Service at Great Ormond Street Hospital.

To be classified as a rare disease in Europe, a condition must affect fewer than five in 10,000 of the general population. At the same time, the sheer number of conditions in this category (estimated to be between 5,000 and 8,000) means that taken collectively, 'rare' diseases are anything but. Young patients are particularly affected, with three quarters of rare diseases affecting children, and 30% of diagnosed children dying before they reach five years old.

While patient numbers may be high, the rarity of individual conditions makes them notoriously hard to diagnose and treat. According to Lucy Jenkins, the interim director of the Regional Genetics Laboratories and consultant clinical scientist at the North East Thames Regional Genetics Service at Great Ormond Street Hospital (GOSH), one of the problems is the multisystemic nature of many such diseases, affecting several parts of the body. This can lead to what has become known as the 'diagnostic odyssey' whereby patients undergo many different tests and hospital visits, passing through the clinics of multiple specialists without necessarily coming any closer to a precise diagnosis.

"Any one specialist may not have a holistic enough view to know what's going on," Jenkins says. "It's not unusual for patients to be seen, for example, by an ophthalmologist and then an endocrinologist and neither of them are looking at the whole picture, just at their particular area of expertise. And of course because of their rarity, it may be that a doctor has never come across a particular condition before and wouldn't know what to look for."

Genetic component
For this reason, GOSH takes a multidisciplinary approach with different specialists assembling their knowledge of the symptoms to try to construct an overall picture of the condition. Given that 80% of rare diseases have a genetic component, it's no surprise that genetic testing is a key part of the investigative work carried out by the hospital.

Tests undertaken (either directly within the GOSH lab or by partner services) can range from single-gene analysis, to exome sequencing and whole genome sequencing, with each bringing its own distinct benefits, costs and challenges.
"We used to be able to look just at one gene and then another and another, taking quite a bit of time," Jenkins says. "But this means there's a long delay for the patient to get a result and get a diagnosis, and it's expensive as well.

"Where a patient is presenting with a set of features that aren't easily defined, it can be very difficult to know what particular genes need to be tested. Even if you have perhaps a more defined clinical presentation like epilepsy, there are many different genetic causes, so the analysis of multiple genes in one go becomes an advantage.

"This can be in smaller panels of maybe ten to 20 genes, or even larger ones of 100 genes plus, so you're looking at a lot of genes associated with that particular presentation rather than just one at a time. That means you are really looking at all of the genetic material in trying to decide what might be going on."

While covering more genetic information might seem automatically advantageous, there can also be degree to which less is more. Whole exome sequencing, by focusing only upon the regions of DNA that code proteins, allows scientists to note variations within this coding region and determine what impact they might be having on the patient. By contrast, whole genome sequencing - which takes in the entirety of a person's DNA - means that unless a variation is already known to scientists, the huge volume of information available may serve only to confuse.

"By looking at the whole genome, we're looking not just at the parts of the genome we know about that encode proteins, but at absolutely everything and unfortunately there are big gaps in our knowledge at the moment," Jenkins says. "You're looking at a lot more DNA sequence in the first place, and when you bear in mind that only perhaps 3% of the human genome is recognised as coding, if we find something in the other 97%, it can be very difficult to know what the significance of that is.

"We all have a lot of differences in our DNA, so if you did carry out a whole genome sequence you might find - in anybody that you sequence - hundreds of thousands of variants from what may be perceived as the normal. So the challenge is picking out which of those might actually be significant and which are just neutral variations: the sort of things that determine how tall you are, or what colour your hair is."

Whichever method of testing is used, the results then feed into the hospital's diagnosis and care of the patient deciding whether further investigations need to be made, what the best treatment options are and whether testing should be offered to family members. For this reason, Jenkins and her team need to hold a strong degree of certainty that the mutation they have identified really does cause the condition for which they say it does, and is not a misleading neutral variation. It's here that the multidisciplinary nature of GOSH's approach is so important.

"Clinical scientists in our team can identify changes in the DNA and to some extent say whether they think that is going to be a cause for disease," Jenkins says. "Of course, what we can't always know is whether that gene is relevant to the way the patient is presenting. And so we have scientists and doctors meeting to discuss the results, with us knowing about the more technical aspects and the effect it might have on DNA and proteins in the body, but the clinician of course knowing exactly what symptoms the patient has, and whether what we have found might fit with this."

One of the most interesting aspects of the work Jenkins's team carries out is instances when the genetic variants identified don't match a patient's symptoms quite as expected, or where mutations of the same gene can result in varied clinical presentations within different patients.

Blurred boundaries
Take for example GOSH's investigation into genes that cause immunodeficiency and their separate - at least at first - genetic studies of patients with early onset inflammatory bowel disease and gastrointestinal problems.

"Just a few years ago those two areas were seen as very different clinically, and yet because we're analysing lots of genes in one go, we're now seeing that a gene that we previously thought was just involved in immunodeficiency can actually cause a presentation that impacts the gastrointestinal system," Jenkins says. "The boundaries are a lot more blurred than we perhaps gave them credit for before. Now that we're aware there is obviously this cross talk between the two body systems, it's not as clear what is causing disease."

While rare diseases remain the focus for Jenkins's team, finding out the genetic root of these conditions may also shed light on diseases perceived to be more common, by increasing scientists' knowledge of different body systems and their interaction with each other. At the same time, however, Jenkins cautions that genetic testing has a long way to go before it can reveal such links with any degree of certainty.

"There are companies who will offer to screen your entire genome and give you a detailed picture of your risk of heart disease and various other common conditions, and I really don't think that the technology and the understanding that we have right now backs that up," Jenkins says. Whether such knowledge is achievable in the future - and whether it is ultimately desirable to patients - remains to be seen, but in the work that GOSH and other genetics labs are doing, a target of a different kind is in hand.

Launched in 2012 and currently under way is the NHS England's 100,000 Genomes Project, which will sequence 100,000 whole genomes from NHS patients by 2017. With a focus on rare disease and cancer patients, the project aims to grow a better understanding of the genetic factors that contribute to different diseases. For the rare disease segment of the project, around 50,000 genomes will be gathered; one taken from each patient and two from their closest blood relatives (for example, their parents), to help scientists pinpoint the cause of the condition.

The role of the genetics labs such as the one at GOSH, will be to take the information found within the project and confirm the results in an accredited setting. Jenkins's team will then endeavour to take these findings into practice so that they are reflected in a clinical outcome for patients.

This way of working is already familiar to the lab. Often, patients who have participated in studies in independent research labs and have uncovered the genetic cause of their disease then have no access to further testing for other family members. GOSH is able to provide this testing, confirm the patient's own results and ensure that the research findings are translated into treatment options.
Increased communication

"We're trying to make sure that the pathway is there so it flows through research, through translation, and that it's ultimately validated into care, so that people benefit from it, rather than it just being research that isn't going to have an impact for 20 years or so," Jenkins says.

Increased communication between genetics labs in different hospitals could also drive progress. While GOSH's lab service and others like it are building up their own picture of rare conditions, and classifying variations for comparison with results from their other patients, there is not currently a mechanism for sharing such findings between labs across the NHS. If a framework for this with securely anonymised patient data could be achieved, the pace of diagnosis for rare diseases could potentially be accelerated.

"We are generating huge amounts of genomic information across the UK and internationally as well," Jenkins says. "I think if the scientific community can share their knowledge, then of course when you put it all together it feeds into that bigger picture and it could be used for the benefit of patients."