Creating Made-to-Measure NGS Testing with Multiple Panels

Tell us about your laboratory and the biomarker testing you do...
 

Yvonne Wallis: We both work at the West Midlands Regional Genetics Laboratory, which is the Genomics Laboratory Hub (GLH) for the Central and South GLH Consortium – a part of the NHS Genomics Medicine Service in England. The Central and South GLH serves a population of almost 13 million people across the West Midlands, Oxfordshire and South of England.

James Beasley: When it comes to volume, we currently perform just under 400 cancer panels a month. However, I’m excited to say that we now have funding to double capacity and are ramping up to perform 10,000 panel tests a year.

What is your approach to selecting next-generation sequencing (NGS) panels?
 

Has it developed over time? YW: We’ve been running NGS panels for somatic cancer testing for over a decade, starting with a nine-gene panel, progressively increasing in size – from nine to 28, 172, and now 523 genes in a single panel. However, we now appreciate that having a single capture-based technology does not fit the requirements for all cancer referrals. So we’ve introduced a mid-sized targeted NGS panel using amplicon- based technology for rapid testing, with plans to introduce additional panel tests using this technology to support capacity and contingency. Of course, because DNA-based panels are not always adequate to detect fusions, we also have a 103-gene RNA sequencing panel. It’s remarkable that we started off with a single nine-gene panel and have now moved up to routinely delivering in excess of 500 genes, alongside a large RNA sequencing panel and two medium-sized DNA panels. It’s a really broad church and meets all the criteria for delivery of somatic cancer testing.

Can you please share your results? What do they mean for patients?
 

JB: We initially used a single NGS platform to provide testing for all cancer clinical indications. Diversifying the cancer panel portfolio has resulted in a significant drop in failure rates. Using different chemistries – including both capture-based and amplicon-based approaches – means we are better adapted to processing poorer-quality samples. Rather than reporting failures, we’re able to detect a lot of variants in tissues that may have previously been a struggle to sequence (see Table 1). Overall, our results have changed pretty dramatically just by increasing the variety of available NGS panels.

YW: Turnaround times have also changed. Over the last six months, we’ve been able to reduce the average turnaround time by more than seven days. When we first started the transition to include both mid-sized and large-scale panels, we were running at an average of about 24 days. Now our average is approximately 16 days and we have plans to reduce it to 10–14 days very soon. We can only do that because we have different types of platforms, which lets us work efficiently – well within the timescales professionals need to ensure the best possible outcomes for their patients.

What are your thoughts about the future of biomarker testing in precision oncology?
 

YW: The National Cancer Genomic Test Directory is a large list of tests available to cancer patients across England. Panel tests are becoming an important component of the technology required to facilitate delivery of the directory, because they are able to cope with the increasing number of tests required per patient on limited amounts of tissue. To me, the future will require even more in terms of the number and scope of biomarkers per patient using panel tests. A good example of this would be routine use of panel tests to identify potential clinical trials for all cancer patients. The Test Directory is quite prescriptive as to which essential genetic targets should be tested for particular cancers, which may not be useful for all patients at the end of standard-of-care treatment. I hope that, in the future, the directory will include a pan-cancer clinical trial test entry to address this. Another thing to consider is speed. This field is huge and growing, but turnaround times for tests will still need to be short to offer patients maximum benefit.

JB: My answer would be circulating tumor DNA (ctDNA). Most biomarker testing for solid tumor samples uses biopsy tissue, and not every patient can have their tumor biopsied. As a result, there is a big drive for more regular ctDNA testing within the NHS. It has been an option in the past, but only on a small scale. I think the future will see NGS panels using ctDNA that can step in when a tissue sample is limited or unavailable (or even, for speed, as the first-line test). In light of that, liquid biopsy panel tests can really help with tissue preservation requirements as well as alleviating issues around tumor heterogeneity.

Do you have any words of wisdom for people who believe that “bigger is better” when it comes to panel size?
 

YW: Large panels underpin the potential to deliver highly flexible high-throughput somatic cancer testing. Once it is fully automated (end-to-end) with reporting integration, it will deliver results for cancer patients at any stage along their clinical pathway. However, it’s important to remember that NGS panel testing is rarely plug-and-play. Successful implementation to deliver accurate results requires appropriate infrastructure and expertise to ensure that everything is accurate and interpreted correctly. This is especially important when using large panels that can throw up unknown variants with the potential to be germline. Quality metrics must be carefully considered; it’s crucial to know when you might be looking at a false positive or false negative result. Bioinformatics is critical to the safe analysis of panel testing data. Some panels and platforms come with their own solutions, whereas others require an in-house team of bioinformatics experts to create the pipeline. Even “out of the box” solutions require expertise to fully understand the limitations of the assay and to provide the appropriate limits of detection. To ensure every cancer patient receives the genetic testing they deserve, it’s important to be flexible, keep your options open, and choose the tools that are best for a variety of clinical scenarios. There are circumstances in which large panels are best – and, equally, there are situations that necessitate smaller panels.