Following the launch of the Glasgow Cancer Assays in November, John Windell from The Biomedical Scientist interviewed Dr Susie Cooke, GPOL Deputy Director and Head of Medical Genomics, about the development of the assays and their future. This first appeared on 3 December 2019 in the Biomedical Scientist, copyright IBMS, and is reproduced with kind permission here.
Dr Susie Cooke leads a team that has developed cancer tests, which could lead to a breakthrough in understanding the deadly disease. How do you produce an inexpensive and easy-to-use assay that permits laboratory scientists to extract the maximum information possible from just a small sample of cancerous material? This, says Dr Susie Cooke, Head of Medical Genomics at the Glasgow Precision Oncology Laboratory (GPOL) at the University of Glasgow, was the fundamental challenge for the team that has spent the past five years developing what have become known as Glasgow Cancer Assays. In helping to match patients more closely with suitable emerging therapies and the latest clinical trials, this new suite of tests could help to change the way the disease is treated and managed.
The outlook is good: the assays are already being evaluated by NHS laboratories in England and Scotland. Technically, the major benefit of the tests is that they have been engineered to be suitable for any solid tumour and to set out the genomic events behind the cancer.
“The Glasgow Cancer Assays are large panels,” says Cooke, “so they use standard sequence-capture and next-generation sequencing methodologies. But two key elements set them apart.
“Firstly, the regions of the genome we selected for targeting are based on an objective and rigorous meta-analysis of publicly available genomic data for all solid tumour types, plus literature rescue of biologically important features for rarer tumour types. When you look at other panels there’s surprisingly little consensus on which genes are included – we surveyed eight existing large cancer panels and only 15% of genes tested were common to them. All included genes for which there is no robust evidence of a role in cancer, likely due to the high false positive rate in statistical predictions. This wastes sequencing capacity on irrelevant regions and makes interpretation difficult. We reviewed over 2000 genes and whittled them down to a set of 555 with convincing evidence for a role in solid tumour carcinogenesis.
Secondly, our assays are designed to capture all genomic feature types – coding and non-coding, substitutions and indels, copy number changes, structural variants, fusions and mutational signatures. Our analyses of whole-genome sequences showed that there are as many copy number driver events in cancers as there are mutations, and that inactivating structural variants are a common mechanism for losing function of key tumour suppressor genes, such as RB1 and PTEN. Both of these genes are potential biomarkers and are being investigated in clinical trials, so missing any events in them would risk seriously confounding the trial results.
Our assays are designed to capture all genomic feature types.”
In addition, the great promise of the Glasgow Cancer Assays is that they can be used in routine healthcare settings, potentially making life easier for clinicians and testing laboratories.
“The advantage of having a comprehensive pan-cancer assay is that samples from different tumour types can be processed together.
“This really simplifies delivery for genetics laboratories, as they don’t have to run small numbers of lots of different assays for different indications. Batching in this way also delivers economies of scale and is amenable to automation. For clinicians, the comprehensive nature of the assays means they don’t have to iteratively order lots of different lines of testing. They get all the information up front and this allows forward planning of the patient’s treatment pathway. We’ve seen genomic profiles using these tests that support a choice of first-line approved therapies and a choice of back-up trial options in case first-line treatment fails.”
During the course of its work on the assays, the team focused intently on the everyday challenges of “real-world” oncology.
“We wanted the assays to be suitable for as many patients as possible, and to us that meant working on FFPE material, working on small biopsies from which only small amounts of DNA are available and working on tumour types, such as pancreatic cancer, which have a very high percentage of normal stroma intercalated with the tumour, and therefore require deeper sequencing to get enough data from the tumour cells.”
“We also wanted them to be feasible for resource-limited public healthcare systems, in terms of delivery costs and making sure that the data generated could be processed and stored with minimal specialist infrastructure. Sequence capture technology allows us to meet all of these requirements while still capturing a large enough genomic footprint to deliver virtually all the biologically and clinically relevant information contained in a cancer genome.”
Work like this is hugely time-consuming, demanding a lot of planning, patience and perseverance from everyone involved. Several large hurdles required leaping, says Cooke.
“The main challenge was identifying what genomic features are truly important in cancers. The curation of existing genomic data to identify bona fide cancer genes was labour-intensive but investing up front in understanding the evidence for a gene’s role in cancer does make interpretation easier when you detect variants in those genes. The other aspect that took time was working up a capture strategy for each element and then finding appropriate samples with orthogonal data for testing and validation.”
NHS laboratories are now testing the Glasgow Cancer Assays to see if they can withstand the rigours of frontline work and deliver all they promise. “There are regulatory processes to work through,” says Cooke. “I can’t speak for the decisions the NHS will make based on their evaluations, but from a technology perspective there’s no reason why they couldn’t be available within the next year or two.”
And following this breakthrough, what will GPOL be turning its attention to next?
“We’ll be keeping the assays up to date and also pushing the boundaries even further in terms of what kinds of genomic features we can capture. Our other big focus is software development. We’ve developed a pipeline alongside the assays to really extract the maximum amount of information from the data, so we’ll be looking to get that out as well.”