Nonmetastatic Urothelial Bladder Cancer (UBC) — Shilpa Gupta, MD

DR GUPTA: Hello, I’m Dr Shilpa Gupta. I’m a genitourinary medical oncologist at the Cleveland Clinic. And it is my honor to join Dr Neil Love today to discuss the Year in Review: Clinical Investigator Perspectives on the Most Relevant New Data Sets and Advances in the Bladder Cancer Edition.

Addressing unmet needs in non-metastatic urothelial bladder cancer really has been a key priority over many decades. And we’ve seen so much progress in the last year. And this is really a much-needed effort because only one-third of patients with non-muscle-invasive bladder cancer actually receive intravesical BCG. And even in a country like the US, we are struggling with BCG shortages which affect access to this. And in patients with muscle-invasive bladder cancer, over 50% of patients may not receive curative intent therapy globally.

And there’s a high risk of recurrence in MIBC despite neoadjuvant cisplatin-based chemotherapy and surgery. And 50% of patients are not even eligible for neoadjuvant cisplatin-based chemo. And about one-third of patients refuse neoadjuvant chemotherapy. Furthermore, patients who are not eligible for neoadjuvant chemotherapy which is cisplatin-based do not have any neoadjuvant treatment options to date. And up-front surgery is what they get. And radical cystectomy in general even in the best of hands can have a significant mortality and morbidity and long-term negative impact on sexual function, quality-of-life, and a lot of emotional distress in addition to the risks associated with surgery in the short-term. And development of effective, safe, and durable treatment for non-muscle-invasive bladder cancer and muscle-invasive bladder cancer as well as focusing on bladder sparing treatments is an unmet need.

We’ll start with the non-muscle-invasive bladder cancer. And this space has really exploded in the last 5 years. In the past, this used to be a surgeon’s disease. And now with the advent of novel therapies, it’s become more of a multidisciplinary disease with medical oncologists playing an important role in this. We first had the approval for pembrolizumab in BCG unresponsive non-muscle-invasive bladder cancer back in 2019 based for patients who had carcinoma in situ. And now, we saw the cohort B results from the KEYNOTE-057, pembrolizumab for papillary high-risk NMIBC. And in this, we saw there was around 132 patients enrolled. Median follow-up is still short around 4 months. And we saw that the 12-month complete response rates were 43.5%. And at year 2, it was close to 35% which is quite impressive because we saw that in the other cohort, efficacy does fall over time.

And this is another very novel study in BCG unresponsive patients where they looked at combining immunotherapy, durvalumab, with or without BCG or with external beam radiation therapy which is quite novel for patients with non-muscle-invasive bladder cancer. And this was the GU16-243 ADAPT-BLADDER study led by Dr Hahn and colleagues in which patients got durvalumab every 3 weeks for 8 cycles +/- BCG induction and maintenance as standard. And patients who were in the EBRT arm received concurrent EBRT of 6 Gy x 3 doses in cycle, for 3 days in cycle 1 only. And the primary endpoint was to get a recommended Phase II dose for each regimen. Secondary endpoints were toxicity and complete response rates. And 28 patients were enrolled, 3 patients got durvalumab alone and 13 patients got durvalumab and BCG, 12 patients got durvalumab with radiation. So in this study, we found that the recommended Phase II dose was full doses of all of these modalities, full dose durvalumab, full dose BCG, and radiation 6 Gy. There was 1 Grade 3 treatment-related adverse event of autoimmune hepatitis. And the 3-month complete response rate for all the patients was 64%. And if we break it down to treatment modalities, it was 33% for durvalumab, 85% for durvalumab and BCG, and 50% for durvalumab and EBRT cohorts. And when we look at the 12-month CR, it was 46%, so much lower than the 3-month CR rates, and durvalumab and BCG had 73% CR rates at 12 months, and durvalumab and EBRT had 33% CR rates. So this is hypothesis generating data and more prospective trials are needed. Dr Love?

DR LOVE: So is this now in a Phase III trial? And also, I don’t know that I’ve heard about external beam radiation therapy for non-muscle-invasive bladder before. Is that done?

DR GUPTA: No, it’s not. And that’s why this was a very early trial and really a Phase I setting just to establish the safety for each, for this modality. And we have never done this before.

DR LOVE: That’s really fascinating. Do you know if there’s a Phase III trial?

DR GUPTA: I’m not aware of the Phase III yet. Probably they’ll keep expanding this and can do the Phase II dosing first. And the landscape is changing so much with other intravesical agents, I don’t know where this will fit. But this was certainly interesting to see that it can be done.

DR LOVE: Well the other thing that’s kind of interesting is they’re using BCG even though they’re not responding to BCG.

DR GUPTA: Right. And we have seen other trials like the QUILT study too. And I think that’s a very valid point, especially in countries like the US where there is BCG shortage. And I think the way the mechanism for durvalumab works though is, I don’t know if you really need to combine it with BCG in BCG unresponsive NMIBC. Because we have seen activity of single agent immunotherapy. So I think regardless, by itself, it didn’t look like it was that effective in this.

So we are seeing a lot of novel approaches coming along in NMIBC with novel agents and radiation as we just saw.

And a very novel device that we have seen show efficacy is called the TAR-200 intravesical drug delivery system. And here’s the picture of it. It actually looks like a little pretzel, which enables a sustained release of drug. TAR-200 specifically is for gemcitabine which is placed in the bladder and then it is released in a sustained fashion over time. And it increases the dwell time of the local drug concentration. Now I want to highlight here that intravesical gemcitabine has been used in the bladder for a long time, but that’s not a sustained release formulation. So this trial, the Phase II SunRISe-1 trial, enrolled patients with BCG unresponsive high-risk NMIBC with good performance status with or without papillary disease. They could have T1 or high-grade Ta. And these patients were ineligible for or declined radical cystectomy. And the key eligibility criteria for this cohort was high-risk NMIBC papillary disease only for cohort 4 where they got TAR-200. And cohort 2 is what I mentioned earlier, it was an 80-patient cohort. Primary endpoint for cohort 2 was overall complete response rate. And key secondary endpoints are duration of response, overall response and safety. And for the cohort 4, the primary endpoint is disease-free survival rate. There are also 2 other cohorts in this study in combination with a checkpoint inhibitor.

So we are going to talk about the SunRISe-1 study in the cohort 2. The complete response in patients with high-risk NMIBC with CIS was 76.7% in 23 patients. And this was centrally assessed and the investigator assessed response was 80%. And this was very well tolerated. Mainly, there were just some low-grade 1 or 2 adverse events manageable urinary symptoms. And it was very infrequent to have a Grade 3 or higher adverse event leading to discontinuation. And as we can see here, the treatment duration and response. The median duration has still not been reached. And the median follow-up was 48 weeks in the responders. And the estimates for the duration of response are about 93% at 6 months and 84% at 12 months. And 21 out of 23 responses are still ongoing. So this is a very promising agent. It has a breakthrough designation by the FDA and really being explored in MIBC as well.

I want to highlight here that FGFR is an attractive therapeutic target in NMIBC. And if we look at this picture, we see that while in metastatic disease, FGFR overexpression is seen in less than 20% of patients, it’s highly expressed in non-muscle-invasive disease compared to metastatic disease. And about 30% of patients with high-risk papillary and NMIBC have FGFR alterations. And hence, it’s a rational approach for targeting. So there’s 2 options. There’s the systemic FGFR inhibitor, erdafitinib, which is also approved in the metastatic setting. And then there’s this drug delivery system like the TAR-200. This is called TAR-210. It has erdafitinib as the slow release formulation within the bladder. And it can just stay there for 3 months. So we spare the patient of the systemic toxicities which are quite a lot with the oral erdafitinib.

This is the THOR-2 study cohort 1 where oral erdafitinib versus intravesical chemotherapy was studied in BCG unresponsive high-risk NMIBC patients. So I’ll discuss the study and the toxicities that we saw with erdafitinib. So basically, patients who screened for FGFR mutations or fusions in the tumor tissue were enrolled. And these patients had NMIBC with recurrence after BCG therapy. I want to highlight here that this is the first trial that used the classic definition for BCG unresponsive in the box below. That is patients who had recurrent high-grade Ta or T1 disease within 6 months of completion of adequate BCG therapy which is 5 or more of 6 full doses of initial induction including 1 or more of maintenance course, or 5 or more of 6 full doses at initial induction course plus 2 or more of 6 doses of second induction. In addition to this definition, they also included BCG experienced patients which means patients should have recurrent high-grade Ta or T1 disease within 12 months of completion of BCG therapy. And prior BCG is defined as greater than or equal to 5 of 6 full doses of initial induction or 5 or more of 6 full doses of initial induction plus 1 or more maintenance course in a 6-month period. And half doses or one-third doses were also allowed. And this is more of a real-world setting where everybody doesn’t fit the criteria for the BCG unresponsive based on the timing. So in this cohort 1, we had high-risk NMIBC patients without CIS. And they were randomized to erdafitinib 6 mg a day dose or investigator’s choice of intravesical gemcitabine or intravesical MMC or hyperthermic MMC. I would like to highlight here that in the early phase of this study, higher doses of erdafitinib were tried. They were far too toxic and 6 mg a day was the dose that was chosen. Despite that, the trial was stopped early due to slow accrual and toxicity issues. Primary endpoint was recurrence-free survival.

As we can see here in terms of efficacy, at a median follow-up of 13.4 months, the median RFS was not reached for erdafitinib. And it was 11.6 months for chemotherapy. And at the clinical cutoff, 25 total RFS events had occurred, 11 with erdafitinib and 14 with chemo. When we look at the rates at 6 months, it was 96% of recurrence-free survival rate with erdafitinib compared to 73% with chemo. And at 12 months, it was lower than 6 months but still 77% versus 41%. So this is quite promising, right? But when we look at the safety and even at this low dose of 6 mg, you can see that patients had, majority of patients, over three-quarters had nail toxicities which can become really bad associated with a really bad day-to-day quality-of-life, hyperphosphatemia, eye toxicities which can be very significant, dry mouth, skin toxicity and central serous retinopathy which can occur in up to 39% of patients and Grade 3 or higher in 4%. And this is something very concerning and that’s where the TAR-210 comes into the picture.

And this is the study that looked at that with patients who had archival tissue or urine cell-free DNA tested for NGS. And in this, there was cohort 1 which was high-risk NMIBC who were BCG experienced or unresponsive and not receiving radical cystectomy. But they’re required to have a complete resection on TURBT of all visible disease prior to treatment. And in cohort 3, patients with intermediate-risk NMIBC were enrolled, those with recurrent and history of low-grade only Ta or T1 disease as well. And, again, visible target lesions prior to treatment was allowed. So in dose escalation Bayesian-based design, there was 2 dose levels, TAR-210-B which was the dose level 1 and TAR-210-D dose level 2. And part 2 was the dose expansion. Primary endpoint for this early-phase study is safety.

So we saw that the steady state mean plasma concentrations are 50 times lower with this intravesical delivery system compared to oral erdafitinib of 9 mg daily. There was no hyperphosphatemia seen which is a really big advantage because we have to continuously monitor patients on oral erdafitinib. And this provides sustained erdafitinib release in the urine over 90 days with very low plasma concentrations, as you can see on these graphs of the urine and the plasma.

And when we look at the efficacy in cohort 1, the FGFR altered high-risk NMIBC patients, there were 16 patients enrolled. Primary endpoint was recurrence-free survival, and it was 82% here. And median recurrence-free survival was not met here, as you can see here. Median duration of treatment exposure was 4.3 months in the dosing of TAR-210-D. And median duration of treatment exposure was around 3.7 months in the other dosing. So these are still very few patients, but we are seeing encouraging activity. And in intermediate-risk patients in the cohort 3, we saw complete response rates of 87%. And this is still not a long follow-up, so just hypothesis generating.

And now we’ll talk about the other novel agent antibody drug conjugant, enfortumab vedotin, which is already approved in advanced setting both in refractory setting and now in front-line setting along with pembrolizumab. This study is exploiting its use in non-muscle-invasive setting. Now just for refreshing, it’s an anti-Nectin 4 monoclonal antibody and the MMAE is the active agent and these are linked by a proteus cleavable linker. And when this agent goes inside the cell, there is endocytosis, MMAE is released and causes microtubule disruption and apoptosis.

So this is the study, EV-104, where there’s intravesical use of enfortumab vedotin. So this is not systemic use. Again, the first of its kind study here. So patients who had BCG unresponsive CIS with or without papillary disease either unfit or refusing radical cystectomy, ECOG performance status 0-2. They received induction with intravesical EV, 25 mL weekly for 6 doses and then went into a maintenance phase from months 4 to 12 every month for 9 doses and they followed up. Primary endpoint was safety, tolerability, feasibility, and the PK PD parameters. And secondary endpoint was clinical efficacy. So as we are in this first-in-human study, the dose escalation phase identified the maximum tolerated dose at different levels, so 125 mg, 250, 500, and 750 mg. And the study design was optimized to maximize the intravesical drug concentration and limit urinary urgency with a 25 mL dose volume which is manageable. And approximately 18 patients are planned to be treated across these 4 dosing regimens and as of the data cutoff from last year, 6 patients had been enrolled and received EV at the first 2 dose levels. So the study is still going on and this is very early interim data.

As we can see here, only 6 patients were enrolled. All 6 patients completed the DLT period and there were no DLTs observed at the first 2 dose levels of 125 or 250. And the treatment-related adverse events that we saw in this interim analysis was fatigue, dry eye, and micturition urgency. And this brings to that question, Dr Love, of why is it causing the eye toxicity if it is staying within the bladder? So that’s really interesting and that makes me wonder if it’s because of the pretzel that we are limiting the seepage into the plasma. Fortunately, there were no Grade 3 or higher treatment-related adverse events. There were no treatment-related SAEs and no treatment-related adverse events that led to dose discontinuation or reduction. And when we look at the preliminary efficacy here, patient 3 had persistent disease at 3 months, was allowed to stay on treatment until the 6-month assessment here. This has only 6 patients so it’s really difficult to generate data but we do see that at the time of cutoff, this one patient was also not evaluable, patient number 6. And the patient did complete all planned doses of EV here. And 3 achieved a CR at the time of analysis. So out of 5 evaluable patients, we are seeing CR in 3 patients. Very, very early data, limited follow-up so we’ll have to keep on the lookout for the follow-up results from this study.

So that was what we saw in the non-muscle-invasive bladder cancer and I’m sure we’ll see some follow-up this year.

Now switching gears to muscle-invasive bladder cancer where also a lot of activity is going on. When we look at adjuvant immunotherapy-based trials in high-risk muscle-invasive urothelial cancer defined as if a patient got neoadjuvant cisplatin-based chemo then their final pathology had to have ypT2-T4a or node positive disease. And if they did not receive cisplatin-based chemotherapy, then they had to have pT3-T4a or node positive disease and they could decline adjuvant cisplatin also. So we have these 3 big trials here, the IMvigor010 which is the anti-PD-L1 antibody atezo versus observation, primary endpoint disease-free survival, secondary endpoint overall survival, and other clinical endpoints. This was a flat out negative study. No disease-free survival or OS improvement was seen. Then we have the CheckMate 274 which is with anti-PD-1 nivolumab versus placebo. Primary endpoint, again, DFS, similar patient population. We saw improvement in DFS but we have not seen any overall survival data and it has been over 3 years since it was reported. More recently, we saw the third trial results, the AMBASSADOR study, which is also anti-PD-1 agent pembrolizumab versus observation. There were coprimary endpoints of DFS and OS and again in allcomers as well as PD-L1 positive and negative. We saw at ASCO GU that there was disease-free survival improvement and at the first look of the data, there was no OS improvement.

Now we saw the CheckMate 274 extended follow-up which was presented in 2023. We saw that in intention to treat patients, the median disease-free survival was 22 months with nivolumab versus 10.9 months with placebo. So it was pretty similar to what we saw early on. In PD-L1 positive subgroup, the median DFS was significantly higher at 52.6 months with nivo versus 8.4 months with placebo, hazard ratio of 0.52. I want to highlight here that in the US, this is approved for allcomers but in Europe it is only approved for PD-L1 positive patients because of this as well as unnecessary financial toxicity. And I also want to highlight here that in the subgroup analysis in the forest plot, we saw that patients who derived benefit from adjuvant nivolumab were those who had received neoadjuvant cisplatin-based chemo. As we can see here, hazard ratio was 0.52 for those who received NAC, and it was 0.92 for those who didn’t.

DR LOVE: I was just curious, did they show in 274, did they isolate out the PD-1 negative patients?

DR GUPTA: Yeah, they did. At the AUA, they showed the PD-L1 negative subgroup and it did not show as promising activity as the PD-L1 positive. And we also saw here that only patients with the bladder cancer derived benefit. Patients with upper tract disease, there was absolutely no benefit and it is still approved in those patients here.

And this is the cohort that they presented at the AUA last year focusing only on MIBC. And here, we see that with nivolumab, the median disease-free survival only in the bladder cancer patients was 25.8 months versus 9.4 months with placebo. And that brings us to this big question whether it should be used for all-comers. We know it doesn’t work if patients don’t get neoadjuvant cisplatin-based chemo. It does not work in upper tract. And we’ve seen it work so much better in PD-L1 positive patients. So really, we are left with a very small subset that may benefit and we have not seen any overall survival data to date. So I think the big question is whether, you know, what is the utility of an adjuvant therapy if we don’t have OS benefit? And I’ve seen some press release that we may see the first OS data at EAU so we’ll stay tuned.

DR LOVE: Do you generally in your own practice outside of clinical trial, what do you do in those situations, PD-1 negative, upper tract disease? Do you present it to the patient and if they’re willing to participate, allow them to decide? If they ask you what to do, like, what do you say?

DR GUPTA: So I don’t test for PD-L1 but for upper tract disease, I don’t offer it. I offer platinum-based chemo based on the POUT study where it shows disease-free survival benefit. And I also talk to patients about that in the IMvigor010 study, we saw that the benefit was only in the ctDNA positive patients so we try to enroll them on our ctDNA study that we have right now, IMvigor011. But we want to present the data to the patients that there’s no OS benefit reported and we have all these very effective therapies in front-line setting so, and the patient makes an informed decision.

DR LOVE: In these cases, do you offer ctDNA outside of trial to the patient?

DR GUPTA: Yes. I’m not using that for decision making but we are doing in the adjuvant setting to monitor the ctDNA. It is an approved modality.

DR LOVE: But I mean like, for example, in a PD-1 negative case, for example, would you look at cell-free DNA in addition to that before deciding?

DR GUPTA: Yeah, I do gather all that information and then we discuss with the patient.

DR LOVE: Interesting.

DR GUPTA: And then we serially monitor.

DR LOVE: Please continue.

DR GUPTA: Although per the label, we don’t need to do that.

DR LOVE: Right.

DR GUPTA: But I think it’s good to understand the dynamics.

DR LOVE: I was going to say, very similar things going on with cancer, gastroesophageal cancer where it’s approved regardless of PD-1 but just like here, you don’t see benefit. There, even the NCCN says don’t use it but that’s why I was curious so I guess they haven’t gone that far here.

DR GUPTA: No, yeah not in the US. And in the AMBASSADOR study where pembrolizumab was compared to observation for again high-risk urothelial cancer, it included upper tract as well like the other studies. There was a median DFS of 29 months versus 14 months, hazard ratio of 0.69. And in this study, we saw that they were higher rates of neoadjuvant chemotherapy probably because this is all in the US, right, and there’s higher utilization so it is possible that that’s why the median DFS was higher than what we saw with nivolumab.

We have not seen any overall survival at this first interim analysis. Unlikely for it to become positive later but we’ll stay tuned for that.

And now comes, again, enfortumab vedotin monotherapy in the neoadjuvant setting for patients with cisplatin ineligible muscle-invasive bladder cancer from the EV-103 Cohort H. So EV-103 was this huge multiple cohort study and Cohort H looked at EV alone. I have to highlight here that typically for deeming somebody cisplatin ineligible, we go by the creatinine clearance, less than 60%, hearing loss, heart failure, neuropathy, poor performance status. Here, it was interesting that 50% of patients were deemed ineligible due to creatinine clearance less than 60, but 40% were deemed ineligible because of hearing loss. In our routine practice, that is not that high a number and also when patients have hearing loss and we talk to them about curative approaches, they’re usually okay with getting cisplatin. And I think when you go look for hearing loss, you will find more but typically we are not really doing a hearing test on everybody. So that was interesting and I wanted to point that out.

In this study, the pCR rate — first of all, to highlight the EV monotherapy dose, it was 3 cycles before surgery and that complete response rate was 36.4%. This is something we have also seen with single agent immunotherapy, pembrolizumab and atezolizumab, you know, something like in this range. And 19 out of 22 patients completed all 3 cycles. There were no noted delays to surgery but there was 1 death from acute kidney injury in this small cohort, so something to be cautious about.

And another trial similar in this cohort, EV-103 Cohort L is the perioperative treatment with EV monotherapy. So in this, they got neoadjuvant 3 cycles of EV and then they also get 6 cycles in the adjuvant setting. And primary endpoint is pCR rates. And interestingly, I want to mention here that if let’s say somebody has a complete response to neoadjuvant EV, we really don’t know what is the rationale of giving them EV in the adjuvant setting then but the study, that is how it was designed. And 51 patients were treated and patients who completed all 3 cycles, 42. So not everybody can complete all 3 cycles even so we need to be cautious. And again, the pCR rates were 34% and downstaging rates were 42%. Nothing different from single agent immunotherapy but I want to point out here that the toxicity was quite significant. There were 57% patients who had skin reactions which is expected but there was 1 death from Stevens-Johnson syndrome and there were 33% of patients had significant peripheral neuropathy. Now in this setting, are we overtreating with adjuvant if we don’t need to? I think that is a question the Phase III trials will also not answer for us.

DR LOVE: So you said that the single agent IO neoadjuvantly causes those kinds of pathCR rates?

DR GUPTA: Yes, like the —

DR LOVE: Like 30%, 30%?

DR GUPTA: Yes, right. PURE-01 study with pembrolizumab showed that. ABACUS study showed that, so.

DR LOVE: And the other thing I was curious about is, this death from Stevens-Johnson, globally if you had to come up with a number, I mean, it’s only 51 patients, that’s 2%, but what would you say the number is of a much larger number of patients, severe skin reactions, maybe not completely? What number would you give that?

DR GUPTA: I think it’s less than 6, 7% but in the post registration trials in registration setting in France, they had seen I think 3 deaths from skin reactions in the early EV monotherapy studies although in the front-line EV/pembro study, we did not see any deaths from skin reactions so that was good. I’d say it is still rare but it’s something we have to be very cautious about.

DR LOVE: Well I’m hearing more and more about the skin issues. And one of the things I was just sort of flashing on is how much concern T-DXd with HER2-positive disease people have about interstitial lung disease and here you’re seeing deaths also from toxicity and maybe not as many but they’re not that many — when you think about all the attention T-DXd gets for interstitial lung disease, I wonder if maybe this ought to get more attention. I don’t know.

DR GUPTA: It should. It should. That is, I think it is our responsibility. And the peripheral neuropathy I want to highlight can be very disabling with this because patients under report their symptoms and then you actually make them walk, you’re seeing that they are falling and then you dig into it. So as a practice, I do a full skin exam of patients and make them walk because this drug can be very toxic and especially in this setting where we’re aiming for a cure.

DR LOVE: Sure.

DR GUPTA: And we see these toxicities.

DR LOVE: Okay, please continue.

DR GUPTA: And now we are back to the use of TAR-200 which we first discussed in the NMIBC setting in SunRISe study in muscle-invasive bladder cancer, so really a unique setting the first time used in patients who were unfit or refused curative intent therapy. So this was a Phase I study. It’s called TAR-200-103. It’s a global study. Patients with cT2-cT3bN0M0 urothelial cancer who got 4 consecutive 3-week cycles of TAR-200 over 84 days. Primary endpoint was safety and tolerability and then we had secondary endpoints of complete response, partial response, duration of response, et cetera. And again, showing the mechanism of action here. So out of the 35 enrolled patients, there were some TAR-200 related treatment emergent adverse events which are notably urinary symptoms like dysuria, urinary frequency, and generally it was very well tolerated with only 2 TAR-200 removals unscheduled, that is before 3 months were over. And overall, 11 patients had CR and 3 patients had PR. Median overall survival was 27 months. Median duration of response was 14 months. Overall response rates were around 40%. So this is again hypothesis generating data. Conclusions were that TAR-200 is safe and well tolerated in elderly patients with muscle-invasive cancer who refused surgery or curative intent therapy and it shows some preliminary efficacy. And there are trials going on in NMIBC space with TAR-200 and immunotherapy combinations.

And this is again showing the results from that.

And now in the topic of muscle-invasive bladder cancer organ sparing treatment, this was a pivotal study, the Hoosier Cancer Research Network GU16-257 led by Galsky and colleagues in which they used gemcitabine, cisplatin, and nivolumab and then proceeded for organ preservation when possible. So patients got 4 cycles of neoadjuvant gem/cis/nivo. Primary objectives were clinical complete response rate and really a complicated algorithm to calculate the positive predictive value of a clinical CR for a composite outcome which was 2-year metastasis-free survival in patients who forego immediate cystectomy or less than ypT1N0 in patients electing immediate cystectomy. As we will see here, very few patients elected immediate cystectomy who got clinical CR, only 1 patient elected. So 76 patients were enrolled, 33 achieved a clinical complete response which is 43%, and 32 out of 33 opted to forgo immediate cystectomy as expected. The positive predictive value of a clinical complete response was 0.97, thus it met the coprimary endpoint. This is showing here the metastasis-free landmark survival, overall survival. So this was an important hypothesis generating study and we need more of said studies to offer to our patients.

DR LOVE: Just to clarify. So these patients did not get any kind of radiation therapy then?

DR GUPTA: No. But subsequently, they did. It was not part of the proposed treatment. If they did not achieve a clinical complete response as we see here in 39 patients, they could get non protocol therapy. For example, 4 patients got radiation on that cohort.

DR LOVE: Do you think it makes sense to do consolidation radiation therapy even in patients who have a pathCR?

DR GUPTA: In pathCR, I don’t think it is needed if they have a pathCR but we have another trial through the Alliance cooperative group where patients have that option to spare their bladder after a complete response with chemoradiation or surveillance. If they have DDR alterations, they can go surveillance. Otherwise, they can do chemoradiation or surveillance. So I think in patients who don’t achieve a clinical complete response, if patients refuse this treatment, then we definitely should offer chemoradiation though, not just radiation.

DR LOVE: And also, I guess in this trial or other similar trials, are any of these using cell-free DNA? Because that seems like it might be pretty interesting to look at.

DR GUPTA: I think they are. We’ve not seen the data but I think they’ve collected the blood to see what’s happening.

DR LOVE: Okay, please continue.

DR GUPTA: And this is the individual patient outcomes who achieved clinical CR and did not.

And now this is the study that we saw last year, this came out, perioperative chemoimmunotherapy with durvalumab for muscle-invasive bladder cancer. This was a single arm Phase II trial called the SAKK 06/17. There were 61 patients enrolled with cT2-T4a and less than or equal to N1 urothelial cancer. The important thing here is it included bladder, urethra, or upper tract. So this was unique in that regard. And this was a truly perioperative study. They got neoadjuvant gem/cis/durvalumab for 4 cycles and then adjuvant durvalumab for up to 10 cycles. Primary endpoint was event-free survival at 2 years. I want to highlight here that 94% of patients had bladder. Only very few patients had upper tract disease. The 2-year event-free survival rate was 75.7% in this study and when we look at the pathologic responses, we saw that complete responses were seen in 40% of patients who had clinical T2 and 20% of patients who had cT3 disease and downstaging was seen in majority of patients actually. So that is pretty nice data and I think here also, even if patients got a pathologic complete CR, they still got adjuvant durvalumab, right? Do we really need that? That’s the — everyone got that.

And this is another study which used durvalumab along with radiation followed by adjuvant durvalumab after completion of radiation in patients with localized bladder cancer. This was led by Joshi and colleagues. So patients with T2-T4 N0-N2 M0 bladder cancer patients unresectable or unfit for surgery or cisplatin ineligible were enrolled and cis eligible could enroll as well. And so durvalumab was used for 2 doses plus radiation over 7 weeks and the response was evaluated with biopsy and cystoscopy and the primary endpoint was progression-free survival at 1 year and disease control rate post adjuvant durvalumab which meant up to 12 months of adjuvant durvalumab was given in this.

We saw here that median progression-free survival was 21.8 months. Median overall survival was 30.8 months. It was well tolerated, did not see any new safety signals. And on the left, you can see that disease control rate at 1 year was 72.7% and overall response rate was 68% with CR in 54 and PR in 13% of patients.

So that brings us to the end of these updates. But I think this highlights how much progress is going on in the field in the localized urothelial cancer space and there’s a lot to look forward to this year.

Metastatic Disease — Thomas Powles, MBBS, MRCP, MD

DR POWLES: I’m Tom Powles. Thank you for inviting me. Today, I’m going to talk about a Year in Review in urothelial cancer. I’ve called it “Improving Outcomes” because we’ve had 2 big, randomized trials which have shown a survival benefit and I’m going to talk obviously a lot about those. But there’re some other subtle things around other ADCs and other approaches which I’m going to talk about also.

I think the first place to start is the standard of care as it currently stands is platinum-based chemotherapy followed by maintenance avelumab. That is based on this randomized Phase III JAVELIN trial where patients whose cancers had not progressed after 4 to 6 cycles of chemotherapy, were then randomized to avelumab or best supportive care. This is established and I think is widely used. And there have been a series of updated analyses associated with these. The most recent of which was the updated overall survival signal which shows a consistent and persistent improvement in overall survival, a 25% reduction in the risk of death, about a 50% reduction in the risk of progression. I think it’s likely that somewhere between 40 and 50% of patients who are treated for metastatic urothelial cancer are potentially eligible for maintenance avelumab. That’s a really important issue. I think it’s preferential to giving first, second line therapy. I also think it’s preferential to monotherapy immune checkpoint inhibition. So, this is a Goldilocks effect. It’s not too late. It’s not too early. There’ve been some recent subgroup analysis of this population showing it working across broad subgroups of patients. Its adverse event profile, about 10% of patients getting significant toxicity. But all of these drugs are associated some adverse events.

And then this is very small, but essentially this is the most recent update from ASCO GU just recently. And here you can see outcomes in subgroups of patients. And this is some of the longer-term outcome. And you can see from here, the conclusions are that even in the most recent update, this is working across broad subgroups of patients. It remains a standard of care and these data focusing on patient reported outcomes show that maintenance avelumab is not associated with a major decrease in quality of life. You might say, how can a drug with known side effects not be associated with a decrease in quality of life? And I think partly that’s because the tools we have to assess quality of life are perhaps less good than we’d like them to be. Because, as I’ve said before, about 10% of patients do have life-changing toxicity, significant toxicity with immune checkpoint inhibition. But the tools we currently have perhaps are imperfect to assess that.

Neil, you’ve got a question?

DR LOVE: Yeah, I just wonder if you could break down a little bit in terms of the patients not eligible. I assume a big part of that or the main part of it is there are people who have progressed or not responding to initial chemotherapy? Are there other patients who fit in that category?

DR POWLES: Yeah, Neil, that’s a great question. I think it’s true that all patients should have a go at immune therapy, because we don’t know who the responding patients — and we don’t know the patients who don’t respond are. So, when patients are not able to have a go at immune therapy I think that’s a problem, number 1. Number 2 is the biomarker research has been really, really challenging. It’s not been successful. And we’ve tried PD-L1 as a biomarker. We’ve tried tumor mutational burden. We’ve tried a string of other biomarkers, and I can tell you none of them are really primetime ready. And so, the key question here is, which patients are not able to get immune therapy or this approach? And as you said, it’s actually mainly those that progress on chemotherapy. Many people say to me, hold on a second, Tom, but only 20% of patients progress on immune therapy. And that’s actually not correct. 20% of patients progress on the first CT scan with immune therapy — with chemotherapy. But with chemotherapy, you give 6 cycles and at the first scan, and there’s 10% progress there, and then the second scan. And actually, if you look from the first scan between cycle 3 and cycle 6, about 30% of patients have some progression during that period of time. And that’s how you get closer to that 50%. Of course, 10% of the patients are not well enough to have it because of comorbidities, their performance status deteriorated. Some patients have such a difficult time with chemotherapy, they choose not to have it. And that’s when the numbers go down from what should be 70%. I said, no maybe 50%, but then 10% don’t want it and then 10% can’t have it and 10% aren’t well enough. And that’s when you get the real-world maintenance avelumab data, which is closer to 20 or 30%. So, we’re leaving behind some patients with this approach, but this approach is better than all of the other options we’ve had in front of us until recently. Remember the atezolizumab/chemotherapy combination trial was essentially negative for overall survival, and most of the benefit appeared to be coming from the maintenance phase. And that importantly was also true for the 361 trial. The pembrolizumab plus chemotherapy, where again, it’s a negative trial with no survival benefit and most of the benefit coming from the maintenance phase.

The key question which people ask me is, why is chemotherapy and immune therapy not being additive in bladder cancer when the combination has worked so well in lung cancer and other tumor types. And, you know, Neil, we don’t know the answer to that question, but there’s probably something unique in the tumor microenvironment. We might talk about this at the end. We have an excluded phenotype predominantly, in urothelial cancer. That means the T cells don’t get into the tumor. They’re on the outside of the tumor. And it might be that this phenotype means a combination with chemotherapy is not synergistic.

DR LOVE: So, yeah. I guess that leads into the next paper, which I’m really curious to hear you discuss, particularly in light of what you just said.

DR POWLES: So, this is a really great piece of work from the nivolumab group. And here we’ve combined platinum-based chemotherapy, but it’s cisplatin. It’s not gem/cis and carbo. It’s only gem/cis with nivolumab and we’ve compared it to gem/cis alone. And you can see here the other key part to this trial, again, maintenance nivolumab is given after chemotherapy. And I just showed you a 25% reduction in the risk of death in perhaps 50% of patients associated with that maintenance period. So, anything positive in this trial, from a survival perspective, part of — and PFS — part of that at least is going to be driven by this maintenance period. The question is how much of that is driven by maintenance and how much by the combination? And I hope I am going to explain that to you today. And here you can see a significant overall survival advantage. We didn’t see that in the 2 previous trials with chemotherapy combinations.

This here is a 22% reduction in the risk of death. Maintenance avelumab in unselected patients, assuming about half of them get it, is probably associated with about 12% reduction in the risk of death. I said overall it’s a 25% reduction, but only half the patients get maintenance avelumab. So 25 divided by 2, call it 12. So here, a 12% reduction in the risk of death associated with avelumab and this is at greater than 12% — this is a 22% reduction in the risk of death. So, it’s within the ballpark, all coming from that maintenance period. Can we convince ourselves that the cisplatin and the nivolumab is doing something on top of that maintenance period which we didn’t see with pembrolizumab, and we didn’t see with atezolizumab? And actually, the result is, it’s not crystal clear. It’s basically working across broad groups of sub-patients. And it’s important to remember the cisplatin population have a better quality of life, they have better performance status than the carboplatin patients. So, we’re picking the winners a bit. So, it might be that actually the good patients are doing better with the immune therapy. I understand that. Also, the progression-free survival is significantly better. And again, we don’t know how much of this is coming to the maintenance period.

Neil, I draw your attention, chemotherapy is given for about 6 months, actually given for about 4 months. And here you can see the progression-free survival curve doesn’t go apart until the chemotherapy stops. And so, if that’s the case, those who believe that all the benefit is coming from the maintenance period, we’ll show this curve and say, actually the PFS benefit is only really marked after the chemotherapy stops. So, the benefit here is mainly driven by the maintenance phase. But that’s not entirely true because the response rate and the CR rate for the combination is about 10% higher than for the chemotherapy alone. So, the believers believe that there is an additive effect or small additive effect because the response rate is higher. And those people who doubt this, because the other 2 trials have been negative, say actually the progression-free survival curve shows that most of the benefit is coming from the maintenance period.

And here are the response data. And you can see the responses are a little bit higher. And again, if you are a person who says, listen, I want most patients to get a go at immune therapy, in the chemotherapy era, you might say this is a good way of doing it, because at least using this approach you know everyone’s getting a crack at immune therapy. If you’re a purist, a statistical purist, you’d come back and say, actually overall most of the survival benefit is probably coming from the maintenance period and you don’t need to do this. And I would broadly agree with that. I’ve done a little bit of work to see if we could have seen this result coming. I’ve said before, we’ve had a negative atezolizumab study and a negative pembrolizumab trial in the overall population, gem/cis, gem/carbo. What happens when we look at the atezolizumab and pembrolizumab trial in the cisplatin only population. And here you can see. Actually the atezolizumab has the lowest OS hazard ratio. And the pembrolizumab trial has the lowest PFS hazard ratio. Now, neither of those are statistically significant because that’s not how these trials were designed. But I put it to you that actually the other trial showed very similar effects. So this isn’t a black swan event associated with nivolumab that’s totally unexpected. When we looked at the subsets of the previous trial in this space, we could’ve predicted a well-powered trial looking exclusively at the cisplatin population, probably would have been positive. There’s some preclinical work, suggests that cisplatin is different from carboplatin. But there’s also the other factor, that cisplatin patients have better performance status and they tend to do better than the carboplatin population and that might be driving this.

The challenge again, though, when we look at where this benefit is coming from, is actually from a PFS perspective. The curves for all 3 of these only really go apart in the maintenance period. So for all of the positivity I’ve said, suggesting there may be something additive in this combination with cisplatin alone, actually, at least part of the benefit, if not all of it, is coming from that maintenance phase. And that’s why many people look at the chemo/nivo data — this is platin/nivo data — and say actually most of the benefit comes from maintenance avelumab. I’m going to stick to maintenance avelumab, thank you very much. I think both approaches are very reasonable as things currently stand. When you look at the gem/cis — sorry, the gem/cis immune therapy data, and the gem/carbo immune therapy data, and you look for differences in CR, difference in response, PFS, and OS hazard ratios, actually you can see, in broad terms, there are more similarities. Maybe the cisplatin population is performing a little bit better than the carboplatin population. But overall I think it’s fair to say that there is a degree of antagonism between chemotherapy and immune therapy. That’s probably slightly less marked with cisplatin. At least part, if not all the benefit is coming from the maintenance phase. So you don’t need to give cisplatin and nivolumab. It’s very acceptable to say I’m sticking with avelumab. But if you come back and said this is the approach I’d like to pursue, I think that’s entirely reasonable.

DR LOVE: So trying to digest this. I guess 1 question that comes to my mind is, can you envision the possibility that people, for example, if you think about the JAVELIN model. Can you envision the possibility that patients who progress on chemotherapy might actually benefit from IO in the maintenance setting, or it would be second line therapy depending on how you’re looking at it? You follow what I’m saying?

DR POWLES: Yeah. So, Neil, let me put it to you a different way. When we started developing immune therapy, we were hoping that we would cure a big group of patients with immune therapy alone and we hoped it would be about 30% of patients. As it transpires, it looks like second line single agent immune therapy is too late for most patients. Only about 15% or 20% of the patients get it. Up front immune therapy is not good enough at getting control of disease. We lose patients along the way. And actually, we put some patients in harm's way. Combining chemotherapy and immunotherapy showed some antagonism. And, for those patients who progress on chemotherapy, we rescue very few of those with subsequent immune therapy. It just takes too long for the immune therapy to work. And so we’ve ended with maintenance avelumab which didn’t fulfill our ambition. We wanted to do better than that. And when you look at the survival, for example, the nivolumab trial that just came out and previous studies, particularly the 361 trial in the gem/cis gem/carbo population, overall survival is remained stubborn, at somewhere between a year and a year and a half. Actually when we go back in time, and we go back 5 years, the survival of bladder cancer was still about 12 months. So immune therapy’s probably made some benefit and made a huge benefit for some patients. But when we look at the whole population, survival remains poor. I do kidney cancer too. When we started, the survival of kidney cancer and bladder cancer was about the same. Kidney cancer’s now 5 years. And until recently, bladder cancer was stuck at somewhere between a year and a quarter and a year and a half. And so immune therapy didn’t fulfill our expectations.

DR LOVE: So, just to kind of put this all together, particularly in terms of a practical message here, it sounds like if it weren’t for EV/pembro we would still be in maintenance avelumab, at least from your point of view?

DR POWLES: I think maintenance avelumab for allcomers is entirely reasonable. I think for those who said, look, I like the increased response with nivolumab and cisplatin, I’m totally bought into that. For those who said, look Tom, 20% CR rate, I want that for my patients, I’m totally bought into that. But I don’t think it’s essential. And from an overall survival perspective, the hazard ratio of 0.78, 22% reduction in the risk of death, with avelumab probably be accounting for at least 12% of that, means if there are benefits with cisplatin and nivolumab during that initial chemotherapy phase, those benefits are relatively modest and we saw similar benefits for pembrolizumab and atezolizumab and we didn’t jump all over those.

DR LOVE: All right. Please continue.

DR POWLES: EV-302, so this is a different study. Enfortumab vedotin is an antibody drug conjugate. It’s not platin-based chemotherapy. It’s given as a single agent. It’s almost certainly the most active single agent drug we have in urothelial cancer. It has 40% response rates as a monotherapy in platin-refractory disease. Pembrolizumab is an immune checkpoint inhibitor. We know a lot about pembrolizumab. Enfortumab vedotin targets nectin-4. 98% of urothelial cancers overexpress nectin-4. When you give these 2 drugs together in Phase II data, you can get 7 —or 68% response rates in the front-line setting. Gem/carbo, and it was tried in the gem/carbo population. Probably 40 to 45% responses. So when we saw these 70% responses in Phase II, we thought, wow this looks really impressive. We think we can beat platinum-based chemotherapy, gem/cis or gem/carbo with this new combination. It’s an ambitious trial because we’ve tried for 40 years to beat platinum-based chemotherapy and we’ve not had success. So this is a randomized Phase III in front line patients, gem/cis or gem/carbo eligible patients, with progression-free survival and overall survival as the primary endpoint. EV/pembro continued until progression. It was given on a day 1 and day 8 cycle. The progression-free survival showed a 55% reduction in the risk of progression. I said before, overall survival was stuck at about a year. And here we show in unselected patients, the progression-free survival is a year. And we could also see this curve plateau at about 40%. You can see the control arm, and 30% of patients in the control arm got maintenance avelumab but unfortunately the control arm’s coming down to pretty much 10%. So this is a seismic change in progression-free survival. And then when we move forward and look at overall survival we can see a really big difference, of 53% reduction in the risk of death, a hazard ratio of 0.47. It says here the median’s 31. But if you look, the tail of that curve is pretty unstable and the median, I think, is probably going to end up closer to 35 months, closer to 3 years, whereas the control arm, which actually performed at 16 months. As I said before, that control arm is better than control arms we’ve seen before for gem/cis or gem/carbo. That’s probably being driven by the 30% who got maintenance avelumab. And that’s struggling to keep up at 16 months.

And so essentially this is the new way of looking at urothelial cancer. Patients having an opportunity to get close to 3 years, than 1 year in overall survival, with a 50% increase in survival. It works irrespective of the type of chemotherapy, gem/cis or gem/carbo. It beat both of them easily, and it works irrespective of the PD-L1 biomarker. It works across broad subgroups of patients irrespective of platinum chemotherapy, upper tract disease or lower tract disease, cisplatin eligibility, performance status. The response rate close to 70% with a 30% CR rate. These are data points we’ve not seen before in urothelial cancer. As I said before, the control arm performed really quite well. And it’s really important also to look at the subsequent therapy. And you can see the control arm getting quite a lot of subsequent immune therapy, whereas in the EV/pembro arm, those patients, they did progress, will switch back to platinum-based chemotherapy.

The adverse event profile — the overall survival curve is the most important curve, really, for me. But the adverse event profile is equally important because this is the new standard of care. I think patients — and doctors will want to give this to their patients. But it is associated with different toxicities associated with it compared to platinum-based chemotherapy. My experience of platinum-based chemotherapy, it was difficult to give. Platinum is associated with neutropenia. And you’re going to see her neutropenia, pancytopenia. It’s associated with sepsis. It’s associated with renal toxicity. It’s associated with really bad nausea. You know, we had to go out and develop nausea drugs to give to these patients to control their nausea. It’s associated with fatigue. So that’s why, actually, the adverse events are more common with chemotherapy than EV/pembro. But EV/pembro has different toxicity. It has peripheral neuropathy. Peripheral neuropathy seems to accumulate with cycles. And once you get between 6 and 10 cycles most patients are beginning to get a type of neuropathy. It’s a sensory and a motor neuropathy. I really encourage you to stop giving the drug if that’s the case. Put it on hold. And you’ll find the neuropathy gets slowly better. And you can potentially rechallenge after that. And of course it’s associated with skin toxicity. Nectin-4’s expressed in the skin. And during the first 3 cycles, skin rash is really prominent. We need to be really aware of it. And again, do not treat through skin toxicity ‘cause you get into trouble. But my experience is hold, let the skin toxicity resolve, and then rechallenge at a lower dose and you can keep going on the drug. So really important, as we bring this into the community, that we educate and train around adverse events.

Neil?

DR LOVE: Any issues in terms of sort of figuring out what’s going on with these patients? For example, IOs can cause dermatologic toxicities.

DR POWLES: Yeah, well, it’s a great question and it’s a difficult question to answer, speaking frankly. What I can tell you, and we’ve treated about 20 patients with the combination, what I can tell you is that the skin toxicity with immune, with pembrolizumab as a single agent, tends to be associated with a different type of rash. It tends to be a relatively mild rash. It can have a very bad total body type of rash. Whereas this EV/pembro rash, it tends to, once it becomes significant, to cause skin blistering. If you treat through it the skin can actually blister and indeed you can get more things, Stevens-Johnson type syndrome eventually. So it is actually a distinct rash. I think it does have an immune component to it, because I feel it is more common with a combination, or you need to be more prominent — it’s more prominent with the combination. But you’re absolutely right. There’ll be an EV rash but there’ll still be potentially an immune-related rash that’s different and may not be associated with its occurrence in the first three cycles. My advice to investigators who don’t have, or clinicians who don’t have huge experience of this, is to say err on the side of caution particularly in a rash in the first 3 cycles. Don’t give EV through this rash. Put both drugs on hold. If you’re not sure, give steroids. There’s little harm in doing that. And certainly, if you’re really not sure, it would be a mistake to say I’m definite this is EV, I won’t give steroids. I don’t think steroids are a good treatment for EV-related rash by itself. Our experience with that hasn’t been good. But you won’t do patients harm by giving steroids if you think there’s a small chance this is immune driven by pembrolizumab.

DR LOVE: So, just kind of curious. There was a period of time when EV/pembro was only available to patients who were not platinum eligible or cis-platinum eligible. At that point in time I was saying to myself, I think that’s what I want if I had bladder cancer. I’m curious at what point along the line you started to think, this is going to blow away JAVELIN?

DR POWLES: Well, I think — so my personal experience is I thought it was going to blow away JAVELIN when we took part in the study because we were treating patients and they were going into durable remissions. Many of our patients came off therapy and they continued in those durable remissions. And we — I saw a patient yesterday, a really sweet man and his wife who we treated 3 years ago and they’re now on single agent pembrolizumab. They’ve stopped the enfortumab vedotin. They’ve now stopped the pembro too because they stopped after 2 years. And they are in a durable remission with small lymph node-only disease that’s negative on PET. And so I think — for me I thought it was going to show this — you can never tell in clinical trials. That’s the beauty of big, randomized trials across 100s of sites. But I thought it was going to be transformative. And I think it is transformative and your observation is correct. If you are a patient in the United States right now, with front line urothelial cancer, you’d want to have this regime instead of platinum-based chemotherapy which means maintenance avelumab is deprioritized because you’re never going to get there.

And 1 of the questions I get asked is, are there EV/pembro patients? Are there platinum patients? And are there pembrolizumab monotherapy patients? What I can tell you is the following. There are not that many patients who are eligible for platinum-based chemotherapy that would not be eligible for EV/pembro, because platinum-based chemotherapy has its own challenges. Now you might say, what about those patients with an active skin rash on immune suppressive agents. But I’d come to you and say, yeah, they can have platinum-based chemotherapy but that’s not going to be super easy. But you can’t give those patients maintenance avelumab either. You can’t give someone on active immune therapy active immune suppressive drugs. So, a renal transplant patient, for example, you might say you can’t give them EV/pembro. And then you say, okay we can give them gem/cis or gem/carbo depending on their renal function, probably gem/carbo. But you wouldn’t want to give them pembrolizumab or avelumab either. So the question there is, I think this supersedes gem/cis or gem/carbo. And then the more difficult question that comes from that is, are there single agent pembro patients? And I think the answer to that is, there probably might still be some patients. You know, 2 years after a relapse, with lymph node only disease, and a performance status of 2, not feeling great. Let’s say they’ve got, manage multiple comorbidities and their life expectancy is 2 or 3 years. You might say why not have a go with single agent immune checkpoint inhibition, see if we can get control of disease. And I think that’s probably 5% of patients.

The other bit I think that’s relevant is there were quite a lot of gem/carbo patients before and we’d say to them, look, we’ll give you gem/carbo, about a 50% chance of progressing with problems, about a 20% chance of getting immune therapy. You need the drugs. Your life expectancy either way is probably less than a year, you know, liver mets, lots of problems. And those patients — many of those patients, are thinking the immunity, say, you know what, I don’t know if I want those drugs if my performance status is going to be a lot worse. You’re not going to be able to offer me very much life extension. Without the drugs maybe 6 months, with the drugs maybe 9 months, maybe a year. You can see patients — or I’ve had patients come back and said, you know, they’re 78 years old or 85 years old, and they’ve got — on half a dozen different agents, they said look I’ve had a good time. I don’t want to spend the last 6 months of my life battling through chemotherapy. EV/pembro might be a bit different from that. We know we can get good control of those patients. We might be able to transform those patients, and the data suggests we can transform those patients into more reasonable outcomes. They’re not quite as good as the gem/cis population. So under those circumstances, that conversation might be quite different. So instead of saying, you might not make it to a year, you’re going to spend most of the time on chemotherapy and there’s about a 50% chance of progressing through, we can now say actually I think we probably can get some control. Let’s have a go and see what happens. So I think it’s actually going to expand the proportion of patients. I think it’s going to supersede platinum eligibility. And I don’t think we’re going to be talking about platinum eligibility in the front line space in the near future.

DR LOVE: All right. Please continue.

DR POWLES: I’d like to talk briefly about erdafitinib if I may, because it’s been a big year for erdafitinib. It’s been somewhat eclipsed by the EV/pembro, gem/cis/nivo front line data. The THOR trial was a massive undertaking. It was a huge, randomized trial in patients whose cancers had progressed either after chemotherapy and immune therapy or progressed after prior immune therapy alone.

Essentially, what we showed in the first component of this trial, where we randomized against paclitaxel/vinflunine for those patients who had had 1 or 2 prior lines of therapy, is we showed that erdafitinib, an FGFR inhibitor, in patients whose cancers have FGR alterations — so about 1 in, about 2 in 10 patients, those patients’ FGFR alterations, who’ve previously had chemotherapy and immune therapy, this as a third line agent or a second line after chemotherapy and avelumab, this outperformed significantly from a survival perspective, third line single agent chemotherapy. These data are actually quite impressive, and this is on the left-hand side of this curve. We learn 1 or 2 interesting things from this study. Firstly, we showed that those patients who have FGF alterations actually have low PD-L1 expression. We also learned that the erdafitinib is associated with significant toxicity. Just because it’s a tablet doesn’t make it easy. It causes mucositis and causes nail toxicity.

On the right-hand side was a second cohort of this, which was even more intriguing, because we thought — because patients with FGFR alterations have low PD-L1 expression, we expected those patients who’ve had 1 prior therapy but hadn’t had immune therapy before, we thought those patients, erdafitinib would significantly outperform pembrolizumab. But guess what? Pembrolizumab did perfectly well. Hazard ratio of 1.18. And what does this tell us? Well, the first thing it tells us is erdafitinib is not the only agent in patients who previously had chemotherapy. The erdafitinib should not be sequenced, in my opinion, before immune checkpoint inhibition irrespective of FGFR status and PD-L1 status. And therefore, when you put these 2 trials together, what does it tell us? Well, firstly, erdafitinib is an active agent but pembrolizumab is at least as active, and therefore immune checkpoint inhibition should be prioritized first. And so where does this fit into the patient pathway? Well, in pre-EV/pembro era, first line platinum-based chemotherapy, maintenance avelumab, or second line pembrolizumab/atezolizumab, and then if that patient has an FGFR alteration, I think single agent EV or single agent erdafitinib are both totally reasonable choices. However, EV/pembro, which is the new era, what do we do after that? Well I said before, the commonest treatment regime was gem/cis or gem/carbo. Erdafitinib, we don’t know if it’s as active and we haven’t used it very much post-EV/pembro. And I think we need to look and do trials prospectively in that population to look at its activity. I think it’s a reasonable choice, but I would probably use it currently third line — EV/pembro, platinum therapy, erdafitinib. That’s my idea, although my friend, Yohann Loriot, who’s involved in the erdafitinib program, his preference is to give it directly after EV/pembro. So, as the landscape has changed with enfortumab vedotin and pembrolizumab, so do our subsequent therapies. The level of evidence for the data we have is reduced. Erdafitinib is a reasonable option in this approach. But actually, strong data is in the previous era with platinum-based chemotherapy.

I’ve also got a second slide, if I may, on erdafitinib, and this is the NORSE study that was presented at ASCO. The reason I present this is many in the audience will say, well if EV — when you combine EV with immune checkpoint inhibition, we seem to get an unexpectedly good response. What happens in the front line setting when we combine erdafitinib with immune checkpoint inhibition? And actually, we don’t get quite the same bump. So this trial, FGF altered patients, front line therapy again, and you can see here, in this selected population, erdafitinib plus immune checkpoint inhibition, not really significantly outperforming erdafitinib alone. The response rate, yeah, it’s 10% higher, not 20% higher. The overall survival, 20 months versus 16 months, not a big difference, and not competing with EV/pembro which is more like 30, 35 months. So, in summary, erdafitinib is a single agent. It has activity in the disease. It’s used after immune checkpoint inhibition, not instead of immune checkpoint inhibition. It was previously chemotherapy, erdafitinib — sorry, chemotherapy, avelumab, EV, or erdafitinib in FGFR altered patients. It’s now EV/pembro, platinum-based chemotherapy, erdafitinib, in my opinion.

It's worthwhile talking also about other ADCs because many in the audience will be saying, well EV looks fabulous. What about sacituzumab govitecan? This is a different ADC. It targets TROP2 which is different from nectin-4 obviously, heavily expressed in urothelial cancer, again, and the payload here is SN-38 topoisomerase I, rather than MMAE. So, while it has the same name, an antibody drug conjugate, it’s as different as a bus and a tree from enfortumab vedotin. What we see with sacituzumab govitecan is response rates about 25 to 30% in platinum refractory disease. EV’s response rate more like 40%. Intriguingly, when we combine it in pretreated patients with immune checkpoint inhibition, we don’t see that massive bounce that we saw with EV/pembro.

And so therefore where are we with sacituzumab govitecan? Well, there’s a big study called TROPiCS-04. TROPiCS-04 is a randomized Phase III study. It’s comparing sacituzumab govitecan with single agent taxane therapy essentially in the third line therapy, platinum-based chemotherapy, avelumab, and then sacituzumab or chemotherapy. I think it’s a really promising agent. We’re looking forward to results of that trial. If that trial is positive, I think that sacituzumab will then become a standard therapy, not just in platinum refractory disease, but I think many will want to use it post EV/pembro as well. It’s not without its side effects, particularly hematological side effects. And many people get G-CSF support. It’s also associated with some nausea, some diarrhea and some alopecia.

One of the exciting things, in my opinion, about enfortumab vedotin plus sacituzumab govitecan, when they’re combined together, have response rates of 70%. Now this is really quite high, ‘cause it doesn’t — it looks like these 2 are additive on top of each other and I think ADC doublets, and dare I say it, ADC immune triplets, EV, SG, and pembrolizumab, are going to be a new area of investigation in urothelial cancer. I actually think it’s an incredibly exciting time in urothelial cancer, ‘cause not only have we doubled survival this year, but we’ve also shown that we can combine these drugs together and we may be able to improve their outcomes even further.

And that’s why I wanted to talk about this study here.

There’s another group of ADCs, disitamab vedotin. Disitamab vedotin targets HER2. It has MMAE as the payload. And here you can show about 50% response rates in patients with HER2 expression. Its adverse event profile looks a bit like enfortumab vedotin, more like EV associated with a payload, maybe with less skin toxicity. So this is another really exciting monotherapy drug. And this is currently in randomized Phase III studies with pembrolizumab in this disease. We’ve explored it in patients with low expression. It doesn’t look as active in HER2 low expression. And, we’ve explored it in combination with PD-1 therapy and we can show response rates again of 70% in pretreated patients. So when you think, oh, are we going to plateau in urothelial cancer, because EV/pembro has changed the field for good? I put it to you actually that we’re just going through a period of new, exciting development.

DR LOVE: So, if we could just follow-up with a few questions. KEYNOTE-361? Yeah. Any comments on this study?

DR POWLES: So, I was involved in this study. This is 1 of the studies which I’ve had, I’ve been very lucky and very, in my career, and I’ve been given the opportunity to lead a number of trials including this study. And this study was perhaps the most important negative trial that we’ve done. Everyone thought that chemotherapy plus pembrolizumab would have similar results as it did in lung cancer. But in bladder cancer, instead of hazard ratios of 0.5s and 0.6s, we got hazard ratios of 0.8s and 0.9s. We explain this by saying the combination of chemotherapy with immune, with pembrolizumab during that initial phase was largely antagonistic. We went on in this trial also to look at single agent pembrolizumab. And we showed that pembrolizumab was less good than chemotherapy, getting initial control of disease. And indeed, many patients who got single agent pembrolizumab progressed quickly and never got the opportunity to have chemotherapy and that accounts for the crossing of the curves. We were losing some patients initially.

There’s been some ad hoc analysis from KEYNOTE-361 recently. And I think some of this ad hoc analysis is interesting looking at patients with CR, looking at patients who had just gem/cis, looking at different combinations. And I think the results of this trial show actually, if you look at those patients who do well with immune therapy, who respond, we know they do extremely well. And actually we’ve shown the same thing in the recent 901 trial with gem/cis nivolumab, in that the responders, the duration of response, and the duration of complete response is really quite impressive. And so my takeaway from this is, there are a group of patients who are exquisitely sensitive to chemotherapy and pembrolizumab and chemotherapy and nivolumab together. Unfortunately, we don’t know who they are. And unfortunately they’re relatively infrequent, perhaps 1 in 6 to 1 in 8 patients. And therefore, while this approach you can see, really works in subgroups, and if we did really great biology we might identify who these patients are, enfortumab vedotin and pembrolizumab, in unselected patients, appears to be a more attractive option.

DR LOVE: From what you can tell, do you discern any difference in the outcomes of this trial versus CheckMate 901?

DR POWLES: Yes, a very important difference. CheckMate 901 is statistically positive and therefore is practice changing, whereas this trial, all of these analysis are exploratory, and therefore interesting and worthwhile but not practice changing. So it’s very important to recognize that that gem/cis/nivo is a positive study and therefore can be given as a standard of care, where chemotherapy and pembrolizumab should be considered exploratory. I think, at the same time, once we’ve acknowledged that, which is the most important issue, when we then look at this data, we see more parallels between chemotherapy and pembrolizumab, and chemotherapy and nivolumab. But I wouldn’t want to take that away from the chemo/nivo group who conducted a perspective positive trial because this trial which I was involved with was not that, and therefore it would be unfair of me to say yes, they’re the same.

DR LOVE: So a couple of questions about HER2 positive disease. In terms of what you showed for disitamab, when you talked about HER2 low by IHC, did they also, did you also look at FISH? Were these FISH low?

DR POWLES: Unfortunately, only a very small minority of patients are FISH positive in urothelial cancer.

DR LOVE: That’s what I figured.

DR POWLES: It’s only about 2%.

DR LOVE: Right.

DR POWLES: So we don’t have enough data on that. Back in the day, the UK group did a study called LaMB where we screened 500 patients for HER2 with FISH and we only showed less than 2% were positive. And for that reason we’ve moved on to IHC. And in fact — probably, well, in that trial 60% of patients were HER2-positive.

DR LOVE: Interesting.

DR POWLES: So HER2 positivity is quite high.

DR LOVE: Another question is, would you like to have, based on what you’re seeing, would you like to have disitamab vedotin available right now as monotherapy?

DR POWLES: I think that’s 1 of the most — I could — so the answer to that question will take me either 2 seconds or 2 hours. I’m going to go for the 2 second answer if I may. Because there are so many similarities in terms of the target with EV, until we recognize whether sequencing DV and EV is associated with activity I don’t think we currently can supersede EV. And so I will be giving EV as monotherapy and EV/pembro instead of DV as monotherapy and DV/pembro. And until we show data that DV works after EV, I will be sequencing EV first. So, although it looks the most promising agent out there, we probably need to learn more about sequencing because of the exceptional activity of enfortumab vedotin.

DR LOVE: So this is my other question, which is, any thoughts about T-DXd in urothelial bladder cancer? And would you like to have that available?

DR POWLES: So T-DXd is slightly different because it targets topoisomerase I and not nectin-4. And therefore, it would be reasonable to, with our limited knowledge, and we haven’t sequenced T-DXd after EV/pembro, but it would be reasonable to collect data on that. And there are people who are looking for these pan tumor approvals for this agent. My experience is urothelial cancer drug activity is challenging. We have used T-DXd in urothelial cancer and we haven’t knocked it out the park, speaking frankly. The results look, in my opinion, not dissimilar to sacituzumab govitecan, maybe a little bit higher. Here you can see, there is activity in urothelial cancer but the numbers actually are relatively small, but particularly the 3+ does look promising, perhaps less so for the allcomers at 39%, but still good. So I think this is a promising drug. I think I’d like to see it explored more in urothelial cancer. But, you know, Neil, what I’d really like to see is EV/pembro plus T-DXd as a triplet, and then you might have 80 or 90% response rates, complimentary targets, triplet therapy. ’Cause in the end, you only get 1 really good go at urothelial cancer. And when you have second and third line, we’re following the law of diminishing returns.