Alicia Morgans: Hi, I'm so excited to be here at ESMO 2022 with Professor Nick James who's here to talk about BladderPath, a presentation that he gave at ESMO 2022. Thank you so much for being here.

Nicholas James: Pleasure.

Alicia Morgans: Wonderful. So Professor, can you tell us a little bit about the reason for the study and what you designed?

Nicholas James: Yeah, so for at least a century, bladder cancer's been staged and diagnosed the same way. So you have a bladder, suspected bladder cancer hematuria, you have a rigid cystoscopy and a piecemeal resection with a hot wire basically. And so this gives you histology and it'll tell you whether the tumor is muscle invasive or not.

It's a procedure with really an awful lot of drawbacks. So at least 50% of the time you'll have, if you put radiographic contrast in the bladder, you'll see extra vesical oozing and 5% you get an overt urine leak, so prolonged hospital stay, a lot morbidity and also I suspect actually might be actively spreading the tumor. I mean it's yeah, if you're putting tumor, urine, inflammatory factors blood into a perivesical space, that's kind of quite a nice growth medium.

So I think it may be, it's not a very good diagnostic test. It's often has to be repeated, it's often inaccurate, it's traumatic and it's really not how we diagnose almost anything else. For most cancers we do a biopsy and scans to stage. So what we proposed, there's good data with MR showing that it's very sensitive and specific for muscle invasion. If you see the same sort of things that we now use commonly for prostate.

And so we hypothesized that you could diagnose by doing a biopsy and a scan and then you would expedite times of treatment because you'd make a more informed decision more quickly. You can get the scan organized quickly and therefore your time to definitive treatments should shorten, so that was what we set out to test.

Alicia Morgans: Okay. Well and that makes sense and I think that we've all, anyone who treats bladder cancer is very aware of the risk of complication, but that TURBT is so necessary or at least has been to really ensure that we understand the staging, which can be very complicated in bladder cancer under and over staging quite rampant. So, tell us what you found by using the MR.

Nicholas James: So what we wanted to do, because we were concerned about MR capacity, the initial thing was that the urologist had to assess whether they thought the tumor was probably noninvasive or might be invasive. So we did a feasibility study where we tested the feasibility of using a five point like a scale. And then we only randomized the patients who were possibly muscle invasive, which were about half of the total we estimated, so that meant that we weren't spending a lot of resource scanning patients with TA tumors and stuff like this.

So it turned out that from the feasibility part that the urologists were extremely good at this. So 98% of the patients they thought were probably non muscle invasive, were indeed non muscle invasive. And then of the other 50% of the patients who might be muscle invasive, it was roughly a 50/50 split. So they couldn't really very accurately tell the bad non muscle invasives from the actual muscle invasives, so that was part of the feasibility.

And the second part of the feasibility was that we wanted at least 80% of the patients randomized to MR to get an MR because we thought if we don't hit that sort of number then we can't test whether MR has an impact. And actually it was somewhere in 95% of the patients got the randomized thing. So it's a real world setting. This was in multiple sites across the UK with all the trials and tribulations of the NHS in general and COVID as well for some of the time we were running the trial. So even with all of that we hit that target.

So for the second part of the trial, the efficacy part, what we wanted to show was that pre-trial we'd done an audit of NHS data, which showed that the median time from hematuria clinic presentation to what we defined as the correct treatment for muscle invasive disease, which we said was either the first of chemotherapy, radiotherapy, surgery, or a decision to give palliative care.

We wanted the time for that to come down by 30 days. We expected it to be about a hundred days. And it's pretty consistently a hundred days in the US, across Europe because you've got these multiple handoffs between MDT and re-resection and pathology and booking scans and all the rest of it. So had hypothesized, we have a hundred days on the control arm and 30 days fewer on the experimental arm.

In fact, what we saw was that the median time to correct treatment on the control arm was 98 days, so pretty much exactly what we've predicted it would be even though sites knew that was the outcome for the trial that they were putting patients into because we thought, well they might have actually done it faster knowing that was what we were measuring. But they didn't, they couldn't. And that dropped to 50 days, just over 50 days, so it halved.

It halved for all four treatment options and very strikingly for palliative care, which was a small number of patients, both sides on the control arm, a number of the patients were having a decision to palliative care made on the day they died in one case on the day after they died. Whereas for the experimental arm, the MR was fast tracking people to palliative care if that was the appropriate thing. And they would also of course spared a TURBT.

So right across the spectrum, whether they were getting surgery or palliative care, they were getting treatment much faster. So the second part of the end points was we wanted to ensure that in making one group better, we weren't making the rest of the group worse. So the 75% who were non muscle invasive and the median time to correct treatment for non muscle invasive disease was the same both sides. It was 21 and 23 days, actually slightly shorter with the MR. And if you look at the whole trial population, obviously merging those to pathway two, the experimental pathway did better.

So our conclusion is that you should do a flexible cytoscopy and an MR if you think it might be muscle invasive and then you can make decisions. Now the decision might be that you want to do a TURBT before you do radiotherapy, say. That's fine. Even that was happening, the TURBT didn't count in the trial as being definitive treatment. They were still getting their radiotherapy faster because you knew they were bad patients. So we looked at other things like inappropriate cystectomy. There was no suggestion that we were taking out the wrong bladders because we were staging them wrong, for example.

Alicia Morgans: So this is all so innovative and thought provoking and I think so necessary in urothelial carcinoma, particularly these bladder patients as you're saying. And there are so many paths that they can end up on and so ensuring they get there more quickly, more efficiently is phenomenal.

I wonder from your initial feasibility study and work and the concerns that you had about can we actually do MR for these patients, were the concerns based around the ability to get the scan itself and scheduling? Were they around reading and interpreting the scan? And then once you have the scan, what's the learning curve there? Can radiologists do this easily or does this take some learning?

Nicholas James: So both of those things were concerns. It's just the logistics. So the logistics actually turned out people, if they opened the trial, they sorted the logistics. So the scans. Happened in terms of the interpretation, what we saw was that as sites did more MRs, they were less likely to then do a TURBT just to make sure. They had more confidence in what they saw. So it suggests there was a learning curve and they were more confident in calling muscle invasive disease without going and resecting it as well.

But even if they did go and resect it, they were still getting to the final treatment faster because it allows you to separate out the lower risk non muscle invasives from the high risk non muscle invasives. So if you know that somebody has got a bad looking tumor, you think it might be muscle invasive, turns out to be non muscle invasive, they need a TURBT, they were getting those TURBTs faster than on the standard arm where everybody just got TURBT not risk stratified.

So we're just in the process of pulling in all the scans because we think we can do some quite interesting if you like radio mix around sort of risk factors and we'll have a nice teaching archive of all of the scans and we're going to pull in the tissue samples and stuff as well.

The other thing I should mention actually, so one of the things that people always say about, "Oh you have to do the TURBT to find variant histology." So there's a lot of data show that actually the TURBT is not a very good predictor of variant histology. It very often gets it wrong both ways. So patients called squamous on TURBT are often not squamous when you have the whole cystectomy sample for example.

But also separate work that we did partly with this and partly we've done previous work. So for the initial cohort we consented all the patients having cystoscopy to have a urine sample taken. So these are just patients coming to the hematuria clinic and we did some other hematuria clinic collections as well. We had about 400 patients altogether and we previously shown that we can, with a targeted sequencing panel, that we can identify all the mutations present in the tumor in cell-free urine because the urine's sitting there in the tumor of course.

And so what we wanted to show was that we could use it as a biomarker. So it looks very good as a biomarker, cell-free urine in that you can identify with the panel mutations, you can find who's got the cancer versus who's not with actually quite a hard degree of accuracy. And we've got follow up looking at this as a monitoring test.

But the important point here is that we can, for example, pick up the P53 status of the tumor from the urine without a single cell in it or a whole range of other things. FGFR3, all of the mutations that we think are important we can find in the urine. So the other bit is you can potentially do a liquid biomarker alongside this where you can molecularly profile the tumor on the urine, you can get a biopsy just to confirm that it is cancer just for box ticking and the scan gives you the stage and you can do all of those things minimally invasively and very accurately.

Alicia Morgans: So, that's all really fascinating and I wonder with this cell-free DNA in the urine that you're talking about, how that really compares with urine cytology and it's very clear that it must be much more sensitive because urine cytology is still being used to monitor these patients at least in the US and is not very sensitive, so tell me a little bit more about that.

Nicholas James: So we're find, we explored it in this trial in the context as a potential screening test and it performs very well. The rock curves are very good, it's very both sensitive and specific. We're also separately exploring it as a monitoring tool. So for example, one of my radiotherapy trials, we sample DNA through and out the far side and not surprisingly the mutations disappear in the patients who get a pathological TR.

Also, some other interesting things in there in that not all the mutations disappear. So there's presumably in patients who remain disease free long-term, so there's presumably some of these are background mutations that are precancerous that are in the normal mucosa. And the panel that we've got is a sort of standardized panel. So you don't need to sort bespoke sequence to each tumor in order to do it. You'll get a pretty high coverage rate. So we think it's potentially both very sensitive and very specific and it gives you the molecular profile of the tumor. So very relevant for drugs that target FGFR3 for example. You can pick it up from the urine.

Alicia Morgans: Yes, very, very relevant. And it sounds like it's pretty rapid and the turnaround should be relatively easy. So I look forward to that coming out.

Now just to comment on the MR part of the story though again, I think it's fascinating that we're getting patients to the right treatment more quickly, but I wonder, are you and the team continuing to follow longer-term disease control outcomes to ensure that what we think is the right treatment is actually the right treatment in terms of cancer control?

Nicholas James: So the short answer is yes we are. So this is relatively short follow up because the outcome was time to treatment, which well in some cases was not that quick, but it's still relatively short, so we will report in due course two-year outcomes for all patients.

Alicia Morgans: Okay. I think that will be fascinating. I think all of this is fascinating. If you had to summarize this work, which really, Professor, is quite transformative, what would it be?

Nicholas James: The take home message for us is that for most cancers we do a biopsy to get the histology and we do a scan to get the stage. We don't do that in bladder cancer. We use a hundred year old technology. I think it's time that we ditch that as a staging technique. It's very relevant as a treatment for non muscle invasive bladder cancer. It's not a good treatment for muscle invasive disease. This gives us a way to move away from that and get these patients who have the worst outcomes in bladder cancer, the right treatment faster.

Alicia Morgans: Absolutely. Well I really look forward to your continued progress and I look forward to the dissemination of this technology and technique because I do think it will make a very meaningful difference to patients in particular and to all of us as we speed up the time to getting people to the right treatment, so thank you for your time and your expertise.

Nicholas James: Yeah, thank you. I hope so. It'd be interesting to see how it goes down with urologists, but yeah.

Alicia Morgans: Well we're a multidisciplinary community team and I am sure we can make it work, and thank you again.