The Uromigos Episode 107: Neoadjuvant Chemo/Nivolumab Sparing Cystectomy

Matt Gowski helps The Uromigos take a deeper dive into the phase 2 data of neoadjuvant chemo/nivolumab in muscle-invasive bladder cancer as a potential bladder-sparing strategy. View the ASCO abstract.

Episode Transcript

Tom:
We joined another one of our ASCO 21 podcasts. I’m here with Brian. Of course, Brian was a bit late, although he wasn’t invited, which made it quite complicated, but we almost had our first, but I’ve been put in charge of organization of this particular one, but we are joined by our great friend and colleague Matt Gowski, Matt, we’ve been doing one or two podcasts around ASCO.

And one of the things, the topics that keeps coming up is your study that you’ve led with your collaborators looking at muscle invasive urothelial cancer.

You’ve enrolled a whole load of patients and you’ve given them chemotherapy at gemsis plus nivolumab. And when they have, or if they’ve achieved a CR and about half of them did get a CR you’ve, then basically continued the immune therapy and observed them. And you’ve done something really controversial, which is you’ve not performed radiotherapy or surgery.

You’ve not pursued the definitive curative approach. And there are lots of risks associates with this approach, but actually your results look quite good. Do you want to just go through a bit about why you think this approach is attractive and then talk about the results?

Matt Gowski:
Sure. So, I think, this is not a new concept, of course, in the early end back days, it was appreciated that some patients had a great clinical response to neoadjuvant therapy and some patients declined and cystectomy.

And so, there are data sets of long-term outcomes of those patients suggesting that a subset of patients can be cured with TURBT plus systemic therapy alone. The challenge with those data sets of course, is that there’s a ton of selection bias and their retrospective in all of those concerns. so, this concept has not been tested rigorously in a prospective way. In many studies you can count them on one hand, basically.

This study we designed in 2015 and the main foundation for this study at that time, because this was the time that some of these genomic biomarkers were emerging, that people are saying might help us select which patients might benefit from this approach.

We took a slightly different tact because in my mind, clinical restaging in this situation is a biomarker. You’re defining something, a clinical complete response that needs to be assessed rigorously. It needs to be defined consistently and its performance characteristics need to be assessed.

So basically, you need to develop this as rigorously as you would a genomic biomarker. That was really the concept that we wanted to test. We wanted to give this treatment chemotherapy plus immune checkpoint block eight, define exactly what we meant by clinical restaging, define exactly what we meant by clinical complete responsive, define the performance characteristics of that.

So that’s what we did, and we had the concerns that you raised, but had some rationale for pursuing this suppression

Brian:
And Matt, can you tell the listeners, what did you define as clinical complete response and how did you arrive to that Definition?

Matt Gowski:
All patients after four cycles of Gemsis Nevo underwent an MRI of the bladder if an MRI was contraindicated, then they had a CT, but we have MRI in the majority of patients. We think as, I shouldn’t say we think our radiologists think as do others, that MRI might be a bit better than CT here for local disease.

All patients underwent urine cytology, and all patients underwent cystoscopy with mapping biopsies of the bladder. So that included, clearly, biopsy of anything visible, but if there was nothing visible then biopsy of the prior scar site, plus other sites in the bladder and all of those had to be negative to account for a clinical complete response on the biopsies. There could be low grade papillary disease, but nothing else.

Brian:
And of the patients who weren’t a clinical CR, where did they fail It? Did they have positive cytology, or did you know? [inaudible 00:04:26] I guess what I’m asking? Like disqualified them.

Matt Gowski:
Yeah. Great question. The majority of patients who didn’t have a clinical complete response was based on biopsy proven disease in that could, on that restaging cystoscopy, even if nothing were visible, there was tumor under the microscope. And could …

Brian:
Can you keep going Matt? Sorry.

Matt Gowski:
I was just going to say that could have been carcinoma in situ, but per our criteria that would necessitate proceeding with cystectomy,

Tom:
Let’s just go through some numbers. How many patients did you enroll?

Matt Gowski:
76 patients were enrolled at the time.

Tom:
Sorry, keep going.

Matt Gowski:
At the time of the data lock for ASCO, because we had a delay in enrollment of the last group of patients because of an amendment in COVID. So, at the time of the data lock 64 patients had reached the clinical re-staging time point and 31 of those patients had achieved a clinical complete response. So, 48%.

Brian:
about half

Tom:
And of those patients, how many patients didn’t make it to the first four cycles and how many patients lost their nerve and wanted a Cystectomy?

Matt Gowski:
No patients have not made it to restaging so far. Although we have those additional 12 patients that we haven’t analyzed yet.

Tom:
So, you’ve got a Hundred percent of patients who make it to four cycles and then 50% of those don’t have a CR and they go.

Matt Gowski:
Oh, I’m sorry. I misspoke. So there have been a couple of patients who had to stop treatment early for toxicity reasons. And so, we had one patient who received three cycles of treatment, had some infection related side effects actually went right to cystectomy because their post, they didn’t have a repeat cystoscopy and biopsy, but they had an MRI, which showed a lot of bladder off thickening. So, they went straight to cystectomy and had a pathological complete response, actually just anecdotal. But,

Tom:
And so, you ended up with something like 31 patients that had what you described as a CR and.

Matt Gowski:
That’s right.

Tom:
Of those 31 patients, which is of the 70, let’s call it 40% something on those lines of the other 60% did those, we know that’s Neo has been approach doesn’t seem to be harmful and it’s a standard treatment. So, let’s make the assumption. Those patients ended up doing okay. Is there any evidence that they didn’t do? Those patients that didn’t have the CR that went on to Cystectomy.

Matt Gowski:
Those patients probably have worse outcomes. I think what we have learned so far, which is probably not surprising to you is that clinical complete response is a prognostic biomarker.

But I think the point that you raise is an important one because I didn’t mention this, but the other premise that we had that hasn’t been rigorously tested is that part of the trepidation about this approach is the disconnect between clinical restaging and pathological restaging, albeit inconsistently defined in studies that have assessed the correlation.

There is some disconnect and that’s been reason not to pursue an approach like this. And I think that’s reasonable. However, that really undermines the potential for salvage cystectomy for a delayed cystectomy in patients with locally recurrent disease patients who you might have missed subclinical disease at restaging, it undermines the potential for that approach to salvage outcomes.

And, and so could a risk adapted approach be similar to an upfront cystectomy and everyone approach? I think that concept has not been adequately tested.

Brian:
I have a question of the 31 clinical CR patients? You said, at least by the abstract eight had low recurrence and six underwent cystectomy looks like three had non muscle invasive disease and two or three had more advanced disease.

Were those, obviously patients when they had a clinical CR were followed with cystectomy and other parameters every three months or something, when was that local recurrence or how did they get to cystectomy, I guess is my question.

Matt Gowski:
Our co-primary endpoints were the clinical complete response rate and the ability of clinical complete response to predict “benefit.” And because patients with a clinical complete response could either have immediate cystectomy or not, we had to come up with a composite definition of benefit.

So, benefit was defined as a pathological complete response. If they underwent immediate cystectomy or being two years metastases free, if they didn’t undergo cystectomy. The reason that I’m bringing this up is that our follow-up is not mature enough for that two year.

Brian:
Hmm.

Matt Gowski:
[inaudible 00:09:17] co-primary endpoint. So, the way that we portrayed these data or display these data rather in the presentation at ASCO is in a swimmer’s lane plot. And this sort of gives you a sense for where patients are in follow up when these local recurrences are happening, when the cystectomies are happening and how many patients have not had recurrence and remain with an intact bladder.

So, eyeballing the swimmer’s lane plots we have about let’s see, I think we have about 18 patients beyond a year and 12 of those patients have not had any local recurrence and have an intact bladder.

Tom:
Matt is it fair to say that you, as it currently stands, you don’t feel you’ve put any patient or you, this ethically approved approach, hasn’t put anyone in harm’s way as it currently stands?

Matt Gowski:
I think that’s an important in complex questions. So, I want to answer it directly, but I do want to add that, coming back to this point about delayed cystectomy and patients with local recurrence. The pathology at the time of cystectomy for a delayed cystectomy, we think in part could help inform the answer to your que, right?

Not definitively, but it gives us some flavor. So, in the patients who underwent a delayed cystectomy for a local recurrence, and I should mention no patients had a distant recurrence before a local recurrence in only one of the patients so far in the clinical complete response observation group has had the development of metastatic disease after cystectomy.

In the patients who wonder when a delayed cystectomy, we have three patients who have CIS or lower disease, we have two patients who had T2N0 disease. And we do have one patient who had T4N1 disease. That’s the patient who ultimately developed metastatic disease.

So, my sense is that these delayed cystectomies are salvaging disease at a time where it’s still relatively localized and confined, but ultimately the only way to answer your question for sure, is with a randomized study.

Tom:
Is it fair to say there’s no advantage in delayed cystectomy in no muscle invasive disease? I can see that but here, if that is inevitable, you need a cystectomy at some point, there is no point in a pursuing approach, which do it in three months or do it in 12 months. Is that a reasonable statement?

Matt Gowski:
If you knew that it had to be done? Is that what you’re saying?

Tom:
Yes.

Matt Gowski:
I mean, if you knew a cystectomy needed to be done, then you’re absolutely right. There would be no reason to pursue an approach like this. The reason for pursuing it is that we know that it’s probably not necessary in everyone.

Tom:
And you think that percent is about 33%. Is that at this current time point, is that your gut feeling on this?

Matt Gowski:
I mean, that’s sort of around what we quote for pathological complete response rates, right?

Tom:
Yep.

Brian:
Yeah. How delayed was the cystectomy in those eight patients?

Matt Gowski:
From the time of start of therapy, we have between about six to eight months.

Brian:
Okay. So still relatively early, I know it’s limited follow up, but relatively early, I mean the nodal patient gives you pause, but that patient’s fate may have been sealed even before they went on study. Right? If they had microscopic disease or something,

Matt Gowski:
That is right.

Tom:
My last question, my second last question is when do you think you’ll know whether or not this is safe in the 74 patients? Do you think it’s safe now? I mean clearly the follow up currently is about a year. Probably say that’s not long enough yet and you don’t know enough. When you’ll know enough.

Matt Gowski:
I think we need longer follow up and we need the additional data sets that are testing this approach. So, we need the retain follow up data, which arguably is a different approach because it’s genomically guided but in essence, trying to accomplish the same thing.

So, we need that result. There’s the retain two study that’s chemotherapy plus immune checkpoint blockade like is being explored in our study. We need those results; we need the results of the Alliance study. So, there’s a few data sets that I think combined should provide us with some comfort as to whether or not this should be a consideration for a select group of patients.

Brian:
My last question, but you wouldn’t even with more mature follow up, there needs to be other data/randomized data in order for this to be adopted as a clinical option for patients.

Matt Gowski:
I would certainly feel more comfortable given the stakes here. I don’t think anyone’s arguing that the stakes aren’t high here, but at the same time we know this is accomplishable in a subset of patients. It’s not that we don’t know that, no one’s argued that people have argued that you can’t select those patients appropriately.

Brian:
Right.

Matt Gowski:
and that’s what we’re trying to do.

Brian:
Any final word about biomarkers? Unless you have something to tell.

Tom:
No, I was going to ask the same question about whether or not we can select these patients better at baseline or whether or not we can select them better off the four cycles because you’ve got the CR patients, which is difficult, and I was going to ask whether or not CT-DNA had a role play.

Matt Gowski:
A couple of things. One is that the genomic alterations that are being tested and some of these other studies were integrated into this study prospectively as a secondary endpoint. So, we wanted to validate those biomarkers and we wanted to validate it in the context of clinical response being rigorously defined, because unless you define clinical response, how do you interpret the added value of these biomarkers?

That’s the key secondary endpoint. Some of that data is being presented at ASCO in terms of predicting clinical complete response but that’s not really what we want to do. We want to use those biomarkers to predict long term bladder-intact disease, free survival and those endpoints aren’t mature enough to do those now.

Brian:
Yeah.

Matt Gowski:
In terms of CT-DNA, absolutely. But not in the blood, I think maybe in the urine.

Tom:
Okay. Interesting.

It’s… Maybe that’s not circulating.

Speaker 4:
It depends if you jump up or down enough, I guess.

Brian:
Hey Matt, thanks for joining us on the last minute. I think you might have been on the most Uromigos podcast. If I’m not, I think it’s like four or five now there’s no prize or money going to change hands. I just wanted to note that fact.

Matt Gowski:
But I can guarantee you that’s going on my CV.

Brian:
Definitely not if you want to get promoted,

Tom:
Matt, terrific. Our great study, great work as always.

Brian:
Thanks for joining us. Appreciate it.