Expert discusses novel approach to ctDNA detection in MIBC

In this interview, Richard T. Bryan, MBChB, PhD, MRCS, FacadTM, discusses the publication, “Circulating Tumour DNA Detection By The Urine-Informed Analysis Of Archival Serum Samples From Muscle-Invasive Bladder Cancer Patients,”¹ for which he served as the senior author. Bryan is a professor of urothelial cancer research and the director of the Bladder Cancer Research Center at the University of Birmingham, UK.

Richard T. Bryan, MBChB, PhD, MRCS, FacadTM

Could you discuss the background for this work?

In the last 2 or 3 years we have seen in the bladder cancer setting a number of interesting datasets emerging with regard to circulating tumor DNA and its clinical implications. We know that the presence of circulating tumor DNA at certain clinical time points leads to patients having worse outcomes than patients without circulating tumor DNA being detected. There's an indication that the presence of circulating tumor DNA may indicate response to certain immune checkpoint inhibitors. There is increasing interest in having much more accurate methods for monitoring patients with muscle invasive bladder cancer, [rather] than relying on cross-sectional imaging alone. Again, data would indicate that circulating tumor DNA is a useful modality for identifying patients who may have metastatic disease that is undetectable by conventional cross-sectional imaging. There are a number of current methods and platforms used frequently in clinical studies to detect circulating tumor DNA. Those are from various manufacturers, C2i, Natera, Foundation Medicine, etc. Those platforms generally commence with the deep sequencing of the primary tumor tissue and then the generation of patient specific assays to use for the detection of circulating tumor DNA in a blood plasma sample. So, that's the landscape; ctDNA is important for monitoring patients with muscle invasive bladder cancer. It may indicate response to certain therapeutic approaches and may allow us to monitor this group of patients more accurately.

Now, if we then step back slightly over the course of around the last 8 years, my team and I have devised, developed, validated, and commercialized a DNA-based diagnostic urine test for bladder cancer, which is now available in the marketplace as the GALEAS Bladder Test, manufactured by Nonacus Limited here in the UK. The GALEAS Bladder Urine Test in its final form utilizes the 451 most common single nucleotide variants across 23 genes that are most commonly associated with bladder cancer. One or more of those variants will be found in over 96% of bladder cancers. As a urine test, we can detect new cases of bladder cancer with 90% sensitivity and 90% specificity. Hence, the urine test is now starting to enter clinical use in the UK, being evaluated by a number of NHS units and being evaluated in a large prospective study funded by Cancer Research UK that I'm leading with colleagues from King's College London. Actually, the urine test is now also being reimbursed in the private sector here in the UK by one of our private health care providers, Bupa.

We thought it would be interesting to see if those common mutations that comprise our urine test and the capture-based technology that Nonacus provides as part of that test could also be used to detect circulating tumor DNA. We embarked upon some pilot work. Now, typically, when you're looking for circulating tumor DNA, your starting point, your substrate, if you like, is blood plasma, because blood plasma does not have contaminating germline DNA from white cells. That's what all of the existing commercial based tests rely on, blood plasma. Now, within our very extensive biospecimen collection here at the University of Birmingham, we have very few samples of blood plasma. But what we do have a lot of is serum. In fact, we have well over 1000 serum samples, which we collected from a large prospective study that ran from 2005 to 2011. Those patients were recruited and donated their samples in an era before next generation sequencing. So, we collected serum thinking that we might be using proteomics and mass spectrometry-based approaches for which serum would be really useful.

What were the key points addressed in the publication?

We had serum samples to test the hypothesis that our diagnostic mutation panel for bladder cancer could detect circulating tumor DNA. Our paper describes two novel approaches.

First, [we developed] a methodology for reliably extracting cell-free DNA from small volumes of serum samples. Typically, we were extracting cell-free DNA from about 1 to 1.5 mils of serum. We developed a methodology along with one of our PhD students, Faisal S. BinHumaid, who is the first author on the paper. As part of his PhD, he refined approaches to reliably extract good amounts of cell-free DNA from serum and then size select the shorter fragments of DNA that you typically see as circulating tumor DNA - to enrich what we extracted for the shorter fragments. That's the first interesting part of it. It provides a methodology in which you can extract cell-free DNA from serum samples. That might be relevant to biospecimen collections for any cancer across the world. I'm sure there are millions of serum samples sat in cancer biorepositories around the world, and investigators are perhaps not realizing that they can still do ctDNA research on those samples. I think that is one important aspect of the paper.

The other important aspect is using our panel to then detect the mutant DNA in the cell-free DNA that we extracted from the serum samples. So, rather than deep sequencing the primary tumor to identify the tumor-specific variants and then devise assays for those variants, we just use the same panel. But we used it informed by the data that we'd already obtained from the urine. So, of all the serum samples we analyzed, we already had GALEAS Bladder data from their urine samples. That meant that we knew the tumor-associated variants that we needed to be looking for in the cell-free DNA. That allowed us to detect circulating tumor DNA at a low variant allele frequency of 0.04%. So, a similar sensitivity as one may see with the other commercial platforms that are available, but which require the deep sequencing and the patient-specific assay development.

So, in essence, we think that's a possible way of making circulating tumor DNA detection much more reachable and cost effective for a large number of patients, particularly our UK NHS patients, where the logistics of sending tumor samples and blood samples (predominantly to California for these types of assays to be undertaken) are both complex and expensive. This represents what we would consider a more straightforward and cost-effective approach.

Now, this is the early days. We're not saying that this is a clinic-ready assay. These are our proof of principle data. We're not suggesting that people should use serum instead of plasma. Plasma is a better sample for detecting circulating tumor DNA. Again, we have to replicate this work in a large number of plasma samples, which we're busy doing. But it just shows a different approach for detecting circulating tumor DNA in bladder cancer patients, and it opens up the possibility for other biorepositories that have serum samples to also enter the ctDNA research environment.

What future work is planned based off of this research?

What we've published is proof of principle data. It's in a small number of patients and their samples. It's 40 patients with muscle invasive disease. There's a whole spectrum of muscle invasive disease: patients with organ confined disease, advanced disease, metastatic disease. So, the patient population is very different to the populations you typically see in other ctDNA publications in the bladder cancer setting. Those samples have generally been obtained from patients as part of clinical trials, where there's a very narrow description for trial entry, maybe patients who have got organ confined disease only, or patients who have got advanced disease only. In terms of the next work, we need to start developing patient sample sets from much more clearly defined patient groups and plasma samples so that we can try and mirror the work that others have done, people like Lars Dyrskjøt from Aarhus, Denmark and Thomas Powles in London, UK. We need to try and mirror those important existing datasets, but with our panel-based approach. Those are important next steps for us. That's the work that we've commenced, so we already have access to some plasma samples, but we need to obtain some more. The threshold for what we call a positive or a negative test is likely to be different in plasma-based cell free DNA than it is in serum-based. We assume that it is! So, there's going to have to be more work to be done before this becomes a valid clinical assay for use, but we're well on that way. I'm hopeful that within the next couple of years, we'll be there in the marketplace, alongside some of the other tests with which investigators will be very familiar.

What are the overall take-home messages for urologists?

The key take-home messages are that circulating tumor DNA detection in the bladder cancer setting is valuable. It's important; it opens up new opportunities of how to monitor and treat patients. I think approaches such as ours make ctDNA detection within reach for the majority of clinicians rather than confined to big academic units and clinical trial settings. Our approach makes ctDNA detection potentially applicable to any clinician treating bladder cancer patients. Another important takeaway is if you've got a freezer full of 1000s of serum samples, they're still useful for ctDNA detection research! We'd be absolutely delighted to work with other groups who have serum samples, and we'd be very happy to advise on how we went about it!

Is there anything else that you’d like to add?

We know from other cancer settings, for example, work by the TRACERx team at the Francis Crick Institute in London, [has shown that] across a number of different cancers, there are different molecular subtypes of disease that have different tendencies to shed circulating tumor DNA. While we work on genuinely finding a role for ctDNA detection in frontline clinical settings, we need to understand the molecular subtype context of tumors that do and don't shed circulating tumor DNA. Again, in the bladder cancer setting, work that Dr. Lars Dyrskjøt in Aarhus, Denmark has been at the forefront of, as well as Dr. Thomas Powles in London through the IMvigor010 study. It appears that the basal squamous molecular subtype of muscle invasive bladder cancer is the subtype that is most likely to shed ctDNA into the circulation. That may have implications for how we use ctDNA detection in the clinical setting. There's possibly more to come from that side of things.

Reference

1. BinHumaid FS, Goel A, Gordon NS, et al. Circulating tumour DNA detection by the urine-informed analysis of archival serum samples from muscle-invasive bladder cancer Patients. Eur Urol. 2024;85(5):508-509. doi:10.1016/j.eururo.2024.01.016