Practice-Changing and Emerging Treatment Strategies in Metastatic Urothelial Carcinoma - Guru Sonpavde

November 18, 2021

Guru Sonpavde joins Ashish Kamat sharing takeaways from the 2021 ESMO meeting starting with practice-changing or practice reaffirming studies, followed by other emerging drugs and data. Dr. Sonpavde highlighted the landmark phase 3 EV-301 study that led to the approval of enfortumab vedotin in post-platinum and PD-1 inhibitor-treated patients with metastatic urothelial carcinoma. Among the other studies, he highlights the NORSE phase 2 study in the first-line setting of cisplatin-ineligible patients with activating FGFR3 mutations or fusions and compared patients receiving erdafitinib with erdafitinib plus cetrelimab. He also discusses the phase 2 trial combining pembrolizumab with a novel agent, Ephrin B4-HSA, in patients who were post-platinum with metastatic urothelial carcinoma.  Next, he highlights a randomized phase 3 study done by the French GETUG group, which compared dose-dense MVAC versus gemcitabine-cisplatin, as perioperative chemotherapy for muscle-invasive bladder cancer, VESPER trial. In closing, he highlights the AURA randomized phase 2 study and the CheckMate 901 trial, the first-line phase 3 trial in metastatic patients.


Guru Sonpavde, MD, Bladder Cancer Director, Dana Farber Cancer Center, Faculty, Harvard Medical School, Boston, Massachusetts

Ashish Kamat, MD, MBBS, Professor, Department of Urology, Division of Surgery, University of Texas MD Anderson Cancer Center, President, International Bladder Cancer Group (IBCG), Houston, Texas

Read the Full Video Transcript

Ashish Kamat: Hello, and welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat, Professor of Urologic Oncology and Cancer Research at MD Anderson Cancer Center in Houston, and it is a pleasure to welcome today, a friend, a colleague, and an expert in the field of bladder cancer, Professor Guru Sonpavde, who many of us know, but may not know him in his relatively new role as Director of Bladder Cancer at The Dana-Farber Cancer Institute. Professor Sonpavde is well published, well researched, he attends a lot of conferences, and every year, he kind of puts out this little nugget as to "my top five" or "my top six" abstracts at the particular conference. He's almost gained sort of a following on social media, and I follow him on Twitter, and I look forward to these summaries.

So this year we thought we would ask you, Guru, to come and share those nuggets with our broader audience today. Thank you again for taking the time, and let me hand the stage over to you.

Guru Sonpavde: Thank you, Ashish, very much for the invitation. Also, thank you to UroToday. This is my Urothelial Carcinoma ESMO Congress 2021 highlights in a nutshell. These are my disclosures. So just to go over these highlights one by one, I would like to start with practice-changing or practice reaffirming studies, followed by other emerging drugs and data.

This is the first study I wanted to highlight. This is a retrospective study of EV301. EV301, of course, was a landmark phase 3 study that led to the approval of enfortumab vedotin in post-platinum and PD-1 inhibitor-treated patients with metastatic urothelial carcinoma. The study was positive, there was an improvement in overall survival in all-comers, and this led to the approval earlier this year, the full approval. This retrospective analysis here looks at some difficult-to-treat subsets. This includes, as shown here, these four subgroups of patients that were presented by Dr. Rosenberg at ESMO Congress, which looked at patients who were elderly, 65 or older, patients who had liver metastases, patients who were upper tract primary site of disease, and also patients who had non-response to prior PD-1 or PD-L1 inhibition.

And so, the bottom line is that when you look at all these four subgroups, there was a benefit across these four subgroups for enfortumab vedotin compared with historical chemotherapy, which was either the taxane or vinflunine in this trial. Also, when they looked at toxicities across these four difficult-to-treat subgroups, there was a similar toxicity benefit, actually, for enfortumab vedotin compared with historical chemotherapy, taxane, or vinflunine. So this retrospective analysis is practice reaffirming, and as we know, enfortumab vedotin has had good uptake in the community based on robust findings with improved outcomes, improved survival in patients in the third-line setting, and post-platinum and PD-1 inhibitor setting, of metastatic urothelial carcinoma.

Now, this is the next study I wanted to highlight. This is the NORSE phase 2 study. This is not practice-changing, but it is interesting new data. This was a study done in the first-line setting of cisplatin-ineligible patients with activating FGFR3 mutations or fusions and compared patients receiving erdafitinib with erdafitinib plus cetrelimab. Cetrelimab is a PD-1 inhibitor and erdafitinib, of course, is the FDA-approved FGFR inhibitor approved in post-platinum patients. So this study, again, is in the first-line setting of cisplatin-ineligible patients.

Now, one of the things in the background to remember is patients who respond to erdafitinib with the FGFR3 activating mutations or fusions usually belong to the luminal subgroup of patients, and these are not immune-enriched. So the thinking was that these patients do not respond that well to immune checkpoint inhibition.  However, there is some retrospective evidence, and there is also some translational evidence from the lab, which suggests that erdafitinib might induce immunogenic cell death. So by inducing immunogenic cell death, by increasing the interferon-gamma pathway, it might actually synergize with a PD-1 inhibitor. So that was the thinking behind this study.

So this is a randomized phase 2 study, not a phase 3 study. This analysis here at the ESMO Congress was an interim analysis, with a small number of patients at this point in this study, as you can see, 18 to 19 patients per group. The bottom line is the response rate looked robustly improved with the combination, erdafitinib plus cetrelimab, 68% versus 33%. So that was an interesting signal of activity. Again, this is an interim analysis and we need to await final data. Also, when they looked at toxicities, as you might expect, the combination had more toxicities. You see that the combination group, either erdafitinib and/or cetrelimab were discontinued in 29% of patients, and both erdafitinib and cetrelimab were discontinued in 8% of patients. More toxicities as you might expect with a combination of drugs.  This is an interesting combination, and therefore, the further study appears warranted, and this study, in fact, is ongoing to finish accrual and we will need to wait and see if these data will lead to further phase 3 evaluation. So interesting data, but not practice-changing at this time.

Now, the third study I wanted to highlight was this phase 2 trial, combining pembrolizumab with a novel agent, the Ephrin B4-HSA, which was looked at in patients who were post-platinum with metastatic urothelial carcinoma and they were PD-1 inhibitor-naive, but they had progressive disease. Essentially, this is a new pathway to most of us, so most of us are not using Ephrin inhibitors in the clinic yet, as you know.

The Ephrin B2 pathway is a pathway that is overexpressed in urothelial carcinoma, it's a surface membrane kinase receptor, and this is a pathway that actually is important in cell-cell interaction. And so, when cells come in communication with each other, these pathways can get turned on because of the surface membrane kinases, the Ephrin B2 and its ligand, which can interact and turn on downstream pathways. The soluble Ephrin B4-HSA molecule is a decoy receptor for Ephrin B2. So essentially, it blocks the Ephrin B2 pathway, and it was shown that this was synergistic with PD-1, PD-L1 inhibitors and might improve anti-tumor activity. This was really a non-randomized phase 2 trial done in this post-platinum setting of combining pembrolizumab with this soluble Ephrin B4 decoy molecule.

As you can see here when you look at all comers, there were 70 patients enrolled in this trial, the overall response rate was 37.1%. So again, this compared with historical data here where you see around 20% response rates appeared better. But again, this is not a randomized comparison, of course, with the caveats of historical standard groups. When you look at the Ephrin B2-overexpressing group, this had an even more interesting response rate, 52.2% overall response rate, with a 23.9% complete response rate. The bottom line is that this combination looked very promising in this patient population of post-platinum patients. Also, when they looked at overall survival and progression-free survival, especially in the Ephrin B2-positive patients, you see a really robust overall survival, 21 months, which is really interesting in this setting of post-platinum patients, where if you recall, the median survival with a PD-1 inhibitor alone is less than 1 year.

The toxicities were manageable. As you see here, really, the one toxicity that stood out was hypertension, more hypertension with a combination than has been seen with single-agent PD-1 or PD-L1 inhibition.

The FDA, interestingly, granted this combination breakthrough designation in Ephrin B2-positive patients just earlier this month in September 2021. So a phase 3 trial of this combination apparently is in development to further evaluate this and validate this data.

The next study I wanted to highlight is this randomized phase 3 study done by the French GETUG group, which compared dose-dense MVAC versus GC, gemcitabine-cisplatin, as perioperative chemotherapy for muscle-invasive bladder cancer. This is the so-called VESPER trial, 500 patients were enrolled. Again, although this was labeled as a perioperative trial, the vast majority of patients received neoadjuvant chemotherapy. As you can see here, 88% of patients received neoadjuvant chemotherapy. The primary endpoint was progression-free survival at 3 years.

Now, one of the twists to this study is that the comparison was between four cycles of GC and not four cycles of dose-dense MVAC, but six cycles of dose-dense MVAC, of course, given every 2 weeks with growth factor support. The reason I believe this was done was to keep the duration of treatment the same, approximately 3 months. But as you can see this does confound the interpretation a bit because of the six cycles of dose-dense MVAC, which we are not routinely giving in the neoadjuvant setting of T2 to T4a, N0, N0 muscle-invasive bladder cancer.

This is the patient characteristic slide. As you can see here, evenly balanced patients, GC, and dose-dense MVAC. One of the things I found interesting in this study is that the vast majority of clinical T2 N0, 90% to 95%.  Although in most studies, we see a little bit more of clinical T2 N0, this is somewhat lopsided, in my judgment, to 90%, 95% clinical T2 N0 patients. When you look at the pathological response, you see numerically more pathologic, complete response with dose-dense MVAC, again, six cycles of dose-dense MVAC, not four, 42% versus 36%. And also, patients who had extravesical disease at cystectomy were also a greater number of patients, where GC had extravesical disease, as opposed to the dose-dense MVAC patients.

Now, the primary endpoint is shown here. The figure on the left shows you the PFS curve for all comers. This is perioperative, neoadjuvant, or adjuvant GC versus dose-dense MVAC. This was not statistically significant. This difference, the p-value just fell short, 0.077. But when they looked at a secondary endpoint, which was looking at the neoadjuvant subgroup alone, dose-dense MVAC did beat gemcitabine-cisplatin in the neoadjuvant cohort, with a hazard ratio of 0.70 for progression-free survival. This p-value was significant, again, this was a secondary endpoint. The overall survival is not yet mature and it is still being followed.

The bottom line with the VESPER trial is a lot of discussions have ensued, whether this should mean the standard of care changes to dose-dense MVAC, I think that this issue is confounded by the fact that six cycles of dose-dense MVAC were used, and secondly, because the primary endpoint, which was PFS in all patients, was not met. So having said that, I think that dose-dense MVAC given for four cycles is certainly a reasonable and valid regimen in the neoadjuvant setting in patients who are lymph node-positive disease. Where I might want to reduce the tumor burden and take them to cystectomy, I might want to use six cycles of cisplatin-based chemotherapy, six cycles of dose-dense MVAC would be reasonable in that setting. But at least in my judgment, this has not changed the standard of care to six cycles of dose-dense MVAC. I might also add that the toxicities were more prevalent in the six cycles of the dose-dense MVAC group. Only 60% of patients actually finished six cycles of dose-dense MVAC.

Now, the next study I wanted to highlight was this AURA study, this randomized phase 2 study. This study basically combined avelumab with Gem-Cis or dose-dense MVAC. It is a randomized phase 2 study. Both groups got avelumab, but the backbone was different, GC or dose-dense MVAC, and both groups got four cycles of the chemotherapy. Now, at the end of the day, this is, again, an interim analysis, so basically, what you see here in this cisplatin-based combination group, you see a pathologic remission rate, which was less than T2 disease, as shown here, 57% versus 64%, showing that these regimens are active. Again, this is not a large study at this point. You see a small number of patients in each group. The study continues to accrue. The toxicities were manageable and combinations of PD-1 inhibition plus cisplatin-based chemotherapy have been described in many other phase 2 studies, and these regimens are being evaluated in phase 3 trials, as you know.

Now, one, or a couple, rather, translational studies I wanted to highlight are going to be discussed now. This is the first one with Matt Galsky that looked at whether cisplatin can immune modulate better than carboplatin. This is interesting because, in the first-line IMvigor130 trial, which combined Gem-platinum plus either placebo or atezolizumab, the subset that received cisplatin appeared to have a better benefit. It appeared that maybe cisplatin-based backbones might synergize better, might induce more immunogenic cell death, and synergize better with atezolizumab.

So in fact, this was looked at in the study some more, and when you look at it in PD-L1-high patients, there seems to be a somewhat even more benefit from cisplatin-based chemotherapy combined with atezolizumab. And so this led to further translational studies, and basically, when they looked at peripheral blood mononuclear cells and looked at immune signaling, they actually saw increased TNF alpha signaling in the cisplatin group compared to the carboplatin group. So that's interesting. Again, supporting the hypothesis that cisplatin-based chemotherapy might be more immunogenic and might synergize better with a PD-1 inhibitor.

They also had a separate neoadjuvant cohort, and each of these cohorts had slightly over 100 patients. In the neoadjuvant cohort, when they looked at cisplatin-based chemotherapy patients and the compared pre versus post-tissue, again, there was an elevation in TNF alpha signaling after cisplatin-based neoadjuvant chemotherapy. So again, supporting the hypothesis of cisplatin, perhaps being a better backbone.

I wanted to highlight that there is a study going on, the CheckMate 901 trial, the first-line phase 3 trial in metastatic patients, which is actually dedicated in the substudy in this phase three study, which is dedicated to the combination with a cisplatin-based backbone, so Cis-Gem plus or minus nivolumab. This is the first study, in fact, in which there is a dedicated cisplatin-based backbone in this comparison. Most of the other studies have had this comparison as part of a whole group of platinum-based chemotherapy and not specifically powered for cisplatin.

This is a ctDNA study. This is the next study and the last translation study I wanted to highlight, where we looked at ctDNA using a sensitive amplicon-based profiling platform. We looked at pre and post-immune checkpoint inhibitor ctDNA profiling. This was a sensitive platform, somewhat different because it's an amplicon-based platform. We found that more than 95% of patients had ctDNA alterations found in plasma, so very sensitive, so this might be a useful platform to select patients for a regimen based on profiling noninvasively. And also, we found of interest that there was an association of PIK3CA alterations and BRCA1/BRCA2 alterations with progression of disease, and emergence of these alterations was found at the second time point. I might add that the second time point was at a median of 6 months after the first time point at baseline. So these were not early changes, but somewhat later changes on the immune checkpoint inhibitor.

We also had a control cohort of platinum-based patients of only 6 patients, but a small control cohort, where we actually did not see these new alterations, especially PIK3CA did not emerge in the presence of platinum-based chemotherapy.

So I wanted to quickly summarize these highlights in one slide. I think that the data reaffirms the practice of enfortumab vedotin as therapy for post-platinum and PD-1/L1 inhibitor-treated patients with metastatic disease who are with progressive disease. And also, I think there are promising data for the combination of FGFR plus PD-1 inhibition as first-line therapy in FGFR3-activated tumors. Also, there are promising data now for a novel agent targeting a new target, Ephrin B2, in combination with pembrolizumab in post-platinum patients. Also now, I think that there are supportive data for the role of dose-dense MVAC in the neoadjuvant setting based on the first randomized phase 3 trial that compared perioperative dose-dense MVAC and gemcitabine-cisplatin. I'm not convinced that this actually changes the standard of care, but it is interesting data that should be considered, especially when looking at patients with, perhaps, pelvic lymph node-positive disease that you might want to take to cystectomy given the higher downstaging rate.

I also think that there was interesting data for the combination of dose-dense MVAC plus PD-L1 inhibition, which we have not seen previously. I think the translational studies are continuing to build the momentum for using ctDNA profiling to guide therapy. I think that the rationale for cisplatin-based chemotherapy synergizing better with PD-1 inhibition is building, and I think we need to wait for data from CheckMate 901, which hopefully will confirm these findings. So I think the future is promising and trials need to be considered for every setting of the disease, given that we still have a long way to go to cure most patients. Thank you.

Ashish Kamat: Thank you so much, Guru. This is such a good summary of the Congress, that it's almost like we should have you do this for all Congresses.

Guru Sonpavde:  Thank you.

Ashish Kamat: I see you doing this on social media, and I certainly follow you. I can't think of a better person to summarize the data, other than, obviously, the primary authors of the abstracts themselves. Thank you again so much for taking the time and spending it with us, and hopefully, we will get to see each other in person sometime soon.

Guru Sonpavde: Thanks, Ashish, thanks a lot. Thanks to UroToday.