This week some of the TGMI team were at the European Human Genetics Conference in Milan. Many, many talks were about doing genetic testing using next generation sequencing (NGS). NGS can be used in various different ways to give broad or deep gene testing, ranging from single gene tests to whole genome sequencing. Alongside the science talks, many NGS test providers from all across the globe exhibited their genetic testing services. With such a large number of genetic tests on offer, how do we choose which test to order?
Not all genetic tests are equal
All NGS tests are based on the same starting process. Typically an NGS genetic test starts with a small amount of DNA extracted from cells, blood or saliva. The DNA is cut up into many fragments that cover the genes you want to target. This collection of DNA fragments, with a few modifications, is called a DNA library. The libraries are put through various steps so their DNA code can be read. This is called DNA sequencing. Finally, the DNA sequence is compared to the human reference sequence so any variations can be identified. The individual variations are called ‘variants’, and we all have millions of them.
Although the principles of all NGS genetic tests are pretty similar, every step of the process can be done in slightly different ways. This means every NGS genetic test provider is doing something different to every other provider. This is true, even if they are testing the same genes.For example, last week we discussed how different tests for the BRCA genes can be. And how this can lead to both false positives (finding variants that are not really there) and false negatives (not finding variants that are really there). At one extreme, the 23andMe test only checks for three particular BRCA variants that are only present in specific populations. The standard comprehensive test – which is the one you should order – fully reads the whole BRCA sequence and reliably finds both small and large variants. Many sub-standard tests are available. These do not reliably detect larger variants that delete or duplicate one or more exon (called exon CNVs). It can be very difficult to know how comprehensive a given BRCA test is.
More DNA information is not always better
A clinician suspects their patient has a genetic disease caused by a DNA variant. They have a choice about which test to order. One option would be to order a whole genome sequencing test. This would give them an analysis of all three billion letters of DNA code. But they would need some way to prioritise the millions of variants to find one that has caused their patient’s disease. And although they will have a broad survey of variants across the whole genome, the quality of the test may not be uniform. The depth of information at any particular point in the genome will not always be the same. And it will probably not be as comprehensive as a deep-dive look at one gene, or a panel of specific genes.
Disease information helps you decide which genes to test
How does a clinician decide which variants to prioritise for scrutiny in a whole genome sequencing test? Often they use the features of the person’s disease to decide which data they want to focus on. For example, if their patient has multiple cancers, they will likely focus on the variant data from cancer predisposition genes. If their patient has a hearing problem they will focus on the variant data from hearing loss genes.
But if they already know which genes they want to focus on they also have the choice to test for just that panel of genes. Sometimes this is the best approach, because the gene panel will give you an answer most of the time and a deep, detailed interrogation of the genes is necessary. Gene panel testing is also still cheaper and faster than whole genome sequencing for most health systems, if you include the total costs.
Sometimes this is not the best approach because the disease is not very distinctive and could overlap many different genetic syndromes. For example, intellectual disability is part of hundreds of diverse genetic syndromes. It is better to either look at all the genes (exome sequencing) or the whole genome.
Sometimes no NGS test is appropriate
Occasionally neither broad NGS testing of all genes, nor deep NGS testing of a panel of genes is the best. For some genetic syndromes NGS testing just doesn’t work well yet. This is because some genetic syndromes are not caused by a change in the DNA code of a gene. For example, Beckwith-Wiedemann syndrome is caused by a relative increase of a growth factor called IGF2. This usually results from a change in methylation rather than a change in the DNA code. So you could sequence the whole genome in a patient with Beckwith-Wiedemann syndrome and find nothing. But a methylation test would give you a quick, cheap diagnosis.
Not all genetic test providers are equal
Comprehensive quality management of NGS test processes is essential for the delivery of safe, accurate NGS testing. We have previously blogged about what influences the accuracy of NGS tests and how this can be evaluated. But how does a provider know what quality management standards they should follow? And how do we know if they follow the appropriate quality management standards?
Most countries develop best practice guidance and recommendations to help test providers deliver tests of appropriate quality. And to give healthcare professionals and patients confidence that the provider is delivering quality tests. Test providers can demonstrate they follow best practice guidance by showing they participate in external accreditation practices. For example when ordering a test, look out for whether a provider follows the ISO 15189 requirements for quality and competence within medical laboratories.
Transparency and collaboration
Transparency in genetic medicine is crucial, and we have blogged about it many times. In the context of genetic testing, the terminology returned by a quick web-search for “transparency” rings true: operating in a way that it is easy for others to see what actions are performed; honesty and openness; the availability of full information required for collaboration, cooperation, and collective decision making.
The delivering of a useful genetic test to a patient is a collaboration between the patient, the health professional and the test provider. The patients give clinical and family information that helps the health professional decide the range of diagnostic possibilities. The health professional shares this information with the testing laboratory so they know which data to focus on. Often, the first review of the data does not give a clear answer. The lab then goes back to the health professional with possible causative gene variants. In turn the health professional can do more clinical evaluations which lets them work out which of the gene variants have caused the patient’s disease.
A crucial part of this collaboration is ensuring the patient and health professionals know the limitations of a laboratory’s tests. Our attendance at ESHG this year showed that not all genetic test providers are being as transparent about this as we need them to be. In part this is because the standards, tools and resources needed to evaluate and communicate quality are still being developed. The knock-on effect of this is that health professionals continue to have a hard time making an informed decision about which test to order.