Looking at chromosomes

Here at St. George’s, I’m a clinical scientist, trained in classical cytogenetics. My current position is I’m training officer, so I oversee the training of all the genetic technologists and scientists in the laboratory. The array technology has actually been around since the early ’90s. However, for the arrays to reach the diagnostic laboratory setting and certainly here at St. George’s, we started in around 2011. A number of other laboratories were before that. So generally, within the NHS diagnostic setting, around 2009 onwards. Array CGH– Array Comparative Genome Hybridisation, to give it its full title– actually involves comparing a patient’s DNA sample against a reference DNA sample. And the reference DNA sample is perceived to be a normal sample, without any abnormalities being present.

So when you’re comparing a patient’s DNA against that, what you’re looking for is any losses or gains in the patient that obviously aren’t present there in the reference DNA and therefore would indicate that there’s an imbalance in the patient’s DNA. Comparing array CGH to conventional cytogenetics or G-banding karyotype analysis– essentially, they’re both a whole-genome analysis, which means you’re looking across the whole of the 22 chromosomes, plus the X and Y, unlike targeted molecular testing. So we’re looking across all the DNA of a patient. The difference between array CGH and karyotyping is the resolution at which you’re doing this.

So in terms of G-banded analysis, what we’re able to detect there, in terms of losses, is something in the region of four to five megabases of DNA, which actually can take out 30, 40 genes. In array CGH, we’re down to the gene level and can actually look within certain genes– at just portions of those genes. So there’s a far difference in resolution. So array analysis– we’re probably down to about 100kb– is our baseline detection rate. One of the advantages that CGH offers over conventional karyotyping is the detection of the abnormality rate. Conventional karyotyping will detect abnormalities in around 10% of patients. With array CGH, we can detect abnormalities in 20% or even 25% of patients.

So in terms of offering a diagnosis to the patient, it’s a much more powerful tool. The limitations of array CGH are that it won’t detect balance rearrangements in a patient. It won’t detect triploidy. It’s not very good at detecting mosaicism. Also, what it can detect is very much dependent on the platform that you use. The higher the resolution of the platform, the more that you can detect. You can detect copy number changes down to the gene level. But obviously at nucleotide level, array CGH can’t detect those. One of the other limitations is that we may not be able to interpret what we find on array CGH. Some of the changes, yes, correlate with known pathogenic syndromes.

However, a number of the changes that we may see we can’t interpret. And so, in terms of patient diagnosis, they’re not necessarily any further forward.