Read the Full Video Transcript

Brian C. Baumann: I'll be talking about radiation therapy for local recurrence and for oligometastatic bladder cancer. These are my disclosures. Here's an outline of the talk. First we'll discuss local recurrence and then we'll discuss oligometastatic disease. Those are the key points I want to hit for each topic.

To start off, I think the best way to treat a local recurrence is to prevent one from happening in the first place, and so I'll briefly review the role for adjuvant radiation. The rationale to do adjuvant radiation after cystectomy for patients with locally advanced disease is really the high rates of local failure that have been reported in modern series for patients with T3 and T4 disease and/or node positive disease with rates for, at 5 years, between 20-41%. I think, importantly, in this disease space, chemotherapy has been shown not to reduce the risk of local failures. Neoadjuvant chemo didn't do it in the SWOG 8710 or in the EORTC/MRC trials, and it's unclear if adjuvant immunotherapy has an impact on local failure based on the published data from the trials. Once these failures happen, they're very rarely salvageable and they're associated with a very poor median survival of 9 months. Compared to some of the older data looking at adjuvant radiation, modern technique, IMRT is actually very well tolerated in prospective series, with excellent local control.

These are just some of the, in my opinion, key studies that have reported on adjuvant radiotherapy. I think, arguably, the most important one is the one at the top, which was a randomized phase II trial conducted in Egypt. Patients had to have negative margins, locally advanced disease, it was 120 patients, and they were randomized to adjuvant sandwich chemo, followed by adjuvant radiation, followed by two more cycles of chemo versus just four cycles of chemo. The trial was a positive trial. The primary endpoint was local regional control of 2 years, and the addition of radiation improved that, 96% versus 69% with, again, trends toward improvement in disease-free survival as well, but the trial wasn't powered for that endpoint.

From kind of the same parent trial, there was a comparison of adjuvant radiation alone versus adjuvant chemo alone, again, also with Gem Cis, four times. Adjuvant radiation was found to be comparable to adjuvant chemo for patients with respect to disease-free survival. So in patients who may not be candidates for adjuvant systemic therapy, radiation, at least in this trial, appeared to achieve comparable outcomes. And this is the trial with the majority of urothelial patients, and on subgroup analysis looking at only the urothelial, all of these results held.

There's some retrospective data from the National Cancer Database that our team did that showed an association with improved overall survival for getting adjuvant radiation, and then Valerie Fonteyne's important, very recently published, trial using modern IMRT, it was single arm, but it showed that acute grade 3 GI toxicity, which has always been the concern, is very low, only 6%. And so NCCN guidelines have said that you can consider adjuvant radiation for patients with locally advanced disease, node positive, positive margin.

I think there's an especially strong argument to use adjuvant radiation for squamous cell, and this is prospective data to back that up. This was an unplanned subgroup analysis from the Egyptian trials, 77 patients with pure squamous cell, and looking at the three different adjuvant therapies, adjuvant chemo alone, adjuvant radiation alone, or adjuvant chemo and radiation, giving adjuvant radiation actually improved overall survival versus chemo and that adding chemo to radiation didn't change the outcomes versus radiation alone. The survival curves there are quite stark, with 2-year overall survival at 71 versus 43. So I think there is an argument to be made for doing adjuvant radiation for squamous cell versus systemic therapy, which, so far, appears to be less effective.

So if we're going to treat patients adjuvantly, which sites should we treat? Well, I think it's really important to identify the patterns of failure if you're going to attempt to do that. This was research that we did some time ago when I was at Penn where we mapped the sites of failure in over 400 patients, and you can see that the failures really concentrated in the pelvic lymph nodes, with fewer failures in the cystectomy bed in the presacral region. This kind of led to the development of a consensus contouring atlas through the NRG where we defined the cystectomy bed and how to contour it for the first time in a more rigorous way. But the consensus at that time in 2016 was that you could probably omit the cystectomy bed for margin-negative patients because the existing data at that time showed very low rates of cystectomy bed failures for R0 surgery.

Now, fast-forward to more recently some new data has come out and there's now a new consensus contouring atlas led by Valerie Fonteyne. This shows, really, that the recommendations have changed, and I agree with these new recommendations, to include the cystectomy bed for T3/T4 patients regardless of their margin status. The new data that's come to light is data from Dr. Zaghloul's Egyptian trials showing excellent local control, 96% at 2 years, and a very favorable side effect profile when you include the cystectomy bed for margin-negative patients, because all of the patients in Dr. Zaghloul's trial were margin negative as a requirement. In addition, from Dr. Murthy and others, there's been some more patterns of failure studies that are actually suggesting cystectomy bed failures are more common.

Now, I think that with modern IMRT, and Dr. Ballas knows this from some research that she's done on this topic, we can really spare a lot of the bowel and urinary diversion to a terrific degree. The red here indicates the high-dose region and the green is sort of medium to low dose. So you can see we're treating our target with a lot of sparing. And in the case of the urinary diversion with our physics and dosimetrists, we're able to really carve that out and protect it. So the urinary diversion is getting less than 5 gray, which is really a clinically not meaningful dose of radiation. And it's very important, of course, that we keep the dose to the urinary diversion low to minimize the risk of side effects.

I'm going to shift gears a little bit, talk about salvage radiation once those local recurrences happen. The first topic is, how do you handle local recurrences after primary radiation of the bladder? Well, if the patient recurs in the bladder, I think there's really little to no role for salvage re-radiation for those patients and surgery if it's an option, would be preferred.

I think the situation is a little different if the patients recur in the nodes after primary bladder radiation, and I think as long as the sites of failure are outside of the prior high-dose treatment fields, it's reasonable to consider another round of radiation directed to those nodes. There's no clear right answer as to whether you should treat those nodes with SBRT or do elective nodal IMRT, treating all of the pelvic nodes and then boosting the involved nodes. There is some data that we can perhaps extrapolate from prostate cancer, but that in itself is also retrospective. That basically found lower rates of nodal failure if you took a more elective nodal approach of treating all the nodes and boosting the involved one as opposed to only treating gross disease.

Now, what about in the situation for recurrences after cystectomy? This can be challenging because there's a lot more small bowel in the pelvis once the bladders removed, and this can limit dose escalation. The fundamental problem is that gross disease typically needs about 60+ gray, whereas small bowel tolerance taps out at around 54. So how do we make up the 6 gray difference? Well, if the disease is very close to the bowel, the answer is you often can't.

In this situation for pelvic node recurrences, I still favor treating the pelvic nodes comprehensively, which has been our approach in prostate cancer, and then boosting the involved nodes to a higher dose. I think it's debatable whether or not you should include the cystectomy bed in those node volumes at the same time. But I think if you're able to get away with it, going above 60 gray, it will certainly improve your chances for local control. I think SBRT is also an option in this situation.

This is a case of a patient that I treated recently with locally advanced bladder cancer, they underwent a cystectomy, they received neoadjuvant chemo prior to the cystectomy, and they developed, really, a very advanced left obturator nodal occurrence that extended into the surrounding bone. We did SBRT to this lesion and we were able to treat the tumor to 35 gray, which is close to or, arguably, an ablative dose, with the adjacent bowel getting only 20 gray, which bowel can handle a lot more than that. So this, again, shows how good some of the technology is getting to allow us to treat targets that we couldn't really treat with definitive doses in the past.

Now I'll shift gears again into the oligometastatic disease space. The first topic is metastasis-directed therapy. I think everyone in this room is likely familiar with this trial, SABR-COMET. It was a relatively small trial of 99 patients, randomized phase II trial of best systemic therapy, plus or minus SBRT to all sites of metastatic disease in patients with recurrent oligometastatic disease from solid malignancies. The five-year overall survival was considerably better in the arm receiving SBRT, 42% versus 17%, and that was associated with a 13-month improvement in median overall survival. The survival curves are there on that side.

Now, importantly, there's no subgroup analysis of the urothelial patients who were enrolled on this study. They were eligible to be enrolled, but they were grouped into the other category. But this certainly has gotten a lot of people excited about the potential role for SBRT in oligometastatic bladder cancer. I think the role for it is not really well-defined, but there are some retrospective studies that I think show potential benefit doing either surgery or SBRT.

One of them is this older study from Lehmann from 2009 where the 5-year overall survival was really quite good, 28%. This is in the era before immunotherapy, of course, with the possibility of combining radiation with immunotherapy and there's thought about potential synergistic effects. And again, some smaller retrospective studies that are at least pointing toward potential benefit here. This was a study of 22 patients who were treated to all sites of disease after receiving chemotherapy with at least a PR. Most of these patients had regional nodal disease, but there were patients with distant nodal, some lung mets, and then some other disease. Basically what they found is, compared to historical controls, that 36% of these patients were disease-free at 6 years. So not proof by any means, but certainly some additional kind of corroborating retrospective data that there may be some value in this approach.

This is a larger retrospective series looking at patients who were treated with SBRT. The other trial wasn't strictly speaking SBRT for urothelial cancer. And you could see that local control was excellent, and that's the survival curves there. Progression-free survival, again, I think reasonable compared to historical controls, but you can see that progression-free survival was much better in the cohort, that only had one site of metastatic disease as opposed to two or more. So, similar to what's been seen in other histologies. We can achieve really good local control, but the question is, does that matter if we can't control progression in other sites? And certainly lower-volume metastatic patients seem to do better with this approach.

I want to just focus briefly on, I think, a special case within recurrent oligometastatic setting and that's M1 disease and in particular those most favorable M1a patients who have just para-aortic nodal occurrences. My general approach here, and again, this is really in the absence of any good data, is if they have lower para-aortic recurrences below the renal vein, I'll treat the pelvic and the lower para-aortic region to about 50 gray, and then boost the involved nodes. For the patients with high PA recurrences, I tend to treat those more like a distant met using SBRT to about 35 gray per five, and I'll often consider simultaneously treating the surrounding uninvolved PA region to a dose high enough to adequately treat microscopic disease.

Now, what about kidney toxicity with this approach, especially if we're treating PA nodes that are nestled between the kidneys? Well, again, I'm routinely amazed at the excellent plans that they're able to generate for us and we're able to achieve mean kidney doses that are routinely less than 5 gray. So again, a clinically insignificant dose of radiation to the kidneys that we couldn't have achieved 10 years ago, maybe not even 5 years ago.

The last topic here is radiation of the primary in patients with metastatic disease. I think a lot of the interest for this is coming out of the STAMPEDE trial in prostate cancer, but really, there's no prospective data to my knowledge for doing this for bladder cancer, for oligometastatic disease. There's some retrospective studies that suggest a potential benefit, but there's major selection bias and potential confounding here.

The Seisen study from a number of years ago looked at cystectomy or radiation and found that using the National Cancer Database that there was an association with better overall survival, but there were very few radiation patients included. Only 49. We sort of repeated that analysis only looking at patients treated with combined chemotherapy and radiation and had a much larger cohort of patients, almost 700 total. We found that there was an improvement with an association with better overall survival for the patients who got chemo RT versus chemo alone. And this benefit persisted on propensity-matched analysis and on landmark analysis, and again, the magnitude of the effects remained doing other sensitivity analyses to try to account for selection bias and unmeasured confounders. The median overall survival difference was almost 14 months versus 8.5 months.

In terms of the rationale for a trial in this space, well, I think there's some retrospective data showing potential benefit, but really selection bias could be driving that effect. And importantly, the burden of metastatic disease is not reported in the NCDB. I think the current data is more hypothesis generating and it really needs to be evaluated in a trial. The success of SWOG 1802 asking this question on the prostate cancer role, I think, shows that the topic is of great interest to our community, and I think that such a trial should probably include both cystectomy and radiation, similar to the design of SWOG 1802.

So in conclusion, when on the topic of radiation for local recurrence, adjuvant radiation has been shown in prospective studies to significantly reduce the risk of local failures for patients with locally advanced disease at the time of their surgery, and this appears to be regardless of systemic therapy. I think for patients who have recurred, successful salvage requires a higher dose of radiation to control gross disease than what we need to do in the adjuvant setting, and there, bowel toxicity becomes a greater concern. So this is not the same situation as prostate where salvage radiation is arguably just as good, and potentially better, even, than adjuvant, salvage is much less likely to be successful in bladder cancer. With respect to oligometastatic disease, the data for radiation is limited. It's really just hypothesis generating and I think this is a fertile area for future trials. Thank you all for your time.