Erdafitinib and current treatment landscape for advanced bladder cancer

Share :
Published: 5 Jun 2023
Views: 3268
Dr Yohann Loriot and Prof Arlene Siefker-Radtke

Dr Yohann Loriot (Gustave Roussy, Paris, France) and Prof Arlene Siefker-Radtke (MD Anderson Cancer Center, Houston, USA) discuss the phase 3 THOR study and where it fits in the current treatment landscape for advanced bladder cancer.

The study found that in patients with advanced or metastatic urothelial carcinoma and selected fibroblast growth factor receptor (FGFR) gene alterations, who had received prior treatment with an anti-PD-L1 agent, erdafitinib significantly improved overall survival, progression free survival, and objective response rate in comparison to investigator’s choice of chemotherapy.

Prof Siefker-Radtke explains the tolerability report of erdafitinib in this patient cohort. They also discuss the NORSE, RAGNAR and VESPER trials. Dr Loriot concludes by talking about the other latest developments on the treatment of bladder cancer with Prof Siefker-Radtke.

Phase 3 THOR study
NORSE trial
Testing for EGFR mutations
RAGNAR trial
VESPER trial


This programme is supported by an unrestricted educational grant from Janssen.

YL: Welcome everyone. I’m Dr Loriot from Gustave Roussy Institute in Paris Villejuif, and with me…

ASR: I’m Arlene Siefker-Radtke, a Professor of Genitourinary Medical Oncology at the University of Texas MD Anderson Cancer Center.

YL: So, Arlene, we have a lot of data to discuss today because we have a lot of data in bladder cancer, especially with the FGFR altered metastatic urothelial carcinoma. 

ASR: I completely agree, it really seems to me that it’s erdafitinib’s time to shine for the treatment of urothelial cancers.

YL: So let’s start, maybe, with the THOR, the phase III trial, and then the NORSE trial.

ASR: That would be wonderful.

YL: Okay, so let’s start with THOR. So THOR has a clear rationale – we know that in bladder cancer 15-20% of the patients had FGFR2 or 3 gene alterations and we have a target therapy, erdafitinib, which is a pan-FGFR tyrosine kinase inhibitor. We have shown together, four years ago in a phase II trial, that the drug is active – 40% response rate in this study – but we need to confirm this data and to show that erdafitinib is better than the standard of care. So that’s why we developed this phase III trial called THOR. There are two different cohorts and we reported cohort 1 which compared erdafitinib to chemotherapy in patients with metastatic urothelial carcinoma, FGFR2 or 3 alteration and disease progression on one or two lines of therapy that included immune checkpoint inhibitors. What we have shown is that erdafitinib improved overall survival, improved PFS, improved the overall response rate – 45% versus 11%. So my first reaction to you, Arlene, what are the consequences for you of this THOR trial in the landscape in bladder cancer?

ASR: This trial provides that clear clinical evidence; it’s the first randomised data with FGF-targeted therapy showing its utility in treatment of cancer. When you consider the number of mutations that we had to screen for – 15-20% mutation frequency – we screened thousands of patients to enrol and successfully complete this trial. So, first of all, to complete a trial requiring that much screening takes a lot of work. We’ve seen several other trials where they weren’t able to complete the trial and the trial was closed early. But we were able to complete the screening, not only complete the screening but get patients on treatment. As you mentioned, this was a two-cohort study; the cohort that is being presented at ASCO is the cohort in patients who received prior chemotherapy and immunotherapy for their urothelial cancer and we see clear evidence of clinical benefit. So we now have the level 1 evidence that we need to ensure that erdafitinib remains a standard of care for the treatment of our urothelial cancer patients.

YL: Coming back to this point, we have EV in the landscape because we have level 1 evidence with the EV-301. So one question in daily practice, if we identify a patient with FGFR2 or 3 alterations, they’re in chemotherapy, the patient receives later immunotherapy and then at disease progression what do you do for this patient? Because we have two drugs with level 1 evidence so in daily practice what can we advise to the community?

ASR: I would say we have two great standards; we have level 1 evidence for both. Both enfortumab vedotin and erdafitinib induce responses and that’s important to know because even in a patient for whom we are considering an immune checkpoint inhibitor, if they are having symptomatic visceral metastases, that’s a setting where immunotherapy hasn’t worked as well. So devising a strategy that induces a response can help improve control and improve patient symptoms and that’s where both of these regimens come into play. Enfortumab vedotin, a 40% objective response rate; erdafitinib, a 40% objective response rate based on the phase II data and now we see with the randomised data a 45% objective response. So I think they’re both great choices, perhaps toxicity – if the patient has neuropathy, hard of hearing, then it makes it easier to choose erdafitinib. If patients have an underlying retinopathy then perhaps it makes it a little easier to choose enfortumab vedotin.

YL: So you discussed the immunotherapy. So coming back to the data that were reported in the NORSE trial, so combining erdafitinib and an immune checkpoint inhibitor as a treatment, what are the conclusions of this trial that we developed?

ASR: As you well know, Yohann, the FGF-altered tumours are typically associated with that immunologically cold tissue microenvironment. We call it an immune dessert in many cases. So the goal of the NORSE trial was to combine erdafitinib with the immune checkpoint inhibitor with the goal of potentially enhancing clinical activity. It was done in patients who were cisplatin ineligible and by the typical criteria about 60-70% of patients met that criteria on the basis of poor GFR. Then we saw 20-30% of patients enrolled on the basis of poor performance status. Of course, patients may have met multiple criteria, as they sometimes coexist and go hand in hand. Yet, despite treating this more frail cohort of patients, we did see an objective response rate with erdafitinib of 40%, 45% actually, and in the combination arm 55% when erdafitinib was combined with cetrelimab. This also translated to a progression free survival with erdafitinib of around 5½ months; with the erdafitinib-cetrelimab arm it was around 11 months. Then we saw promising overall survival as well with a median overall survival of 16 months with erdafitinib and 20.8 months with erdafitinib and cetrelimab. So now we’re entering a survival range that has been observed with other targeted strategies like enfortumab vedotin and pembrolizumab.

YL: So it means that we should look for the FGFR status for all patients in the metastatic setting?

ASR: I think we should. Clearly we see evidence of utility, evidence of responses and survival that’s beneficial to the patient population based on your data from the THOR clinical trial. So it’s very important to screen for these mutations as early as one possibly can so that we have the opportunity to give a treatment and not miss an opportunity to give an effective line of therapy.

YL: And what kind of testing will you use? Because sometimes it’s quite challenging to get the tissue and to test, so ctDNA, any value in this setting?

ASR: Well, there is certainly value with that circulating free DNA. I would argue it would have been challenging to do the NORSE trial without it. You draw a vial of blood and usually within a week you get your tissue or you get your FGF mutation result. The challenge with tissue – it takes time to arrange a biopsy, you have to wait for the pathologist to review it then ship the tissue. During the pandemic we saw significant delays in the ability to access tumour tissue.

YL: We discussed [inaudible], THOR and now there’s NORSE, there’s a third, RAGNAR, because we can find FGFR alterations in other solid tumour types. So what about the data from RAGNAR?

ASR: First, I have to point out the obvious Norse connection with Janssen and the Norse god, Thor; erdafitinib named after the goddess of the Earth, Erda, and then we have RAGNAR, which I think is taken from Ragnarok, a big battle that everyone thinks about, and what an epic battle to look for mutations across tumour types. In this tumour agnostic trial a 30% objective response rate for patients who had an FGF alteration present. We saw several tumour types that had an increased response rate – salivary tumours 100%.

YL: Yes, and even in pancreatic cancer a 55% response rate.

ASR: Are there any pancreatic cancer drugs that have such a high objective response rate? I don’t know.

YL: No, I think it’s very striking. So there were 217 patients in this trial and globally, indeed, a 30% response rate is quite striking to me.

ASR: Yes, but out of 200 patients how many did you have to screen for that clinical trial?

YL: Thousands, thousands.

ASR: Thousands. And the presence of these mutations are only present in a small percentage of that tumour population, such a low percent. It would be very challenging to conduct a randomised trial for an FGF pancreatic tumour.

YL: So what about the erdafitinib plan in bladder cancer? So we discussed a lot in the metastatic setting, so, based on this data, what do we think we have to do maybe in earlier stage, so maybe in the perioperative setting or even earlier?

ASR: Absolutely, once it works in the metastatic setting bringing it into earlier disease spaces becomes very relevant. But the challenge is toxicity and if you look at where the FGF mutation is most prevalent, its highest frequency is in the superficial urothelial tumours where over 60% of patients will have an FGF alteration present. But that’s a cohort of patients who have the option of having their bladder out.

YL: And they don’t want any toxicity, right?

ASR: Exactly, to have hand-foot syndrome and dry mouth and mucositis and retinopathy, even though it’s typically reversible, that’s a challenge for patients who can fix those problems by having their bladders removed. 

YL: There is a device now that is an intravesical pretzel, a kind of pretzel.

ASR: Absolutely, this new technology. It’s a drug-eluting pretzel where you can put the drug inside this pretzel. It can be linear, they insert it in the bladder, it curls up, it looks just like a little pretzel and it elutes the drug over time, resulting in increased concentrations inside the bladder. So what a great way of avoiding systemic toxicity and delivering a drug directly where it’s needed.

YL: And for the patients who need oral drugs, what are the most important tips in daily practice for managing the toxicity for the community?

ASR: I think for erdafitinib, monitoring for central serous retinopathy is important. It’s typically reversible but what the current guidelines suggest are baseline ophthalmologic exams, including determining their refraction, what type of glasses they might need, are they 20/20, 20/40 and the like. But we also do optical coherence tomography looking for that retinopathy and then we repeat it. On the trial it was monthly for the first four months. So it’s important to tell your patient if you see blurry lines or blurred vision, I always tell them, ‘Blink your eyes first’ because often they have dry eyes and it could be a film over the cornea. But if they have those blurred or wavy lines, hold the drug, see the ophthalmologist and check for retinopathy. Definitely hold the dose and when the retinopathy resolves most patients can go back on treatment at a dose reduction of the therapy.

YL: And for the nail disorder? Because it could be also a problem for the patients in daily practice. It’s a matter of dose reduction, kind of pose that the tips that we can provide to the community?

ASR: When I think of the toxicity, I think of treatment with erdafitinib as an induction maintenance strategy. We go up to that higher dose, up-titrated to 9mg, and that induces a deeper response. Yet if they have toxicity and we have to dose reduce, which we usually do, we can maintain that response even at an 8mg. My patient who was on treatment for three years was controlled at the 6mg daily dose.

YL: I have one at 4mg.

ASR: So it does have efficacy in controlling those responses and maintaining the responses, even at the lower dose. So I would say don’t feel bad about dose reducing, it’s often needed. I have seen most patients who developed hand-food syndrome, after about 2-2½ months of therapy they need about 2-3 weeks off, allow healing of the hands to occur, it also helps with the nails, and then they can go back on treatment. 

YL: And what can we say about the second cohort, the cohort 2? So what is the expectation or the rationale, just maybe to refresh everyone on this second THOR?

ASR: Well there were two cohorts to the THOR clinical trial. We wanted to determine whether erdafitinib compared to chemotherapy in patients who had prior immunotherapy and then out of concern that the FGF tumour, or the FGF altered tumour, was immunologically cold we also have a cohort of patients who had prior chemotherapy but had not received an immune checkpoint inhibitor. Unfortunately, we don’t have those results to share yet.

YL: So maybe an upcoming meeting.

ASR: Maybe at an upcoming meeting we’ll have something more to share.

YL: So, Arlene, there was also a very interesting abstract on the VESPER trial, this French trial that compared Gemzar cisplatin versus dose-dense MVAC in both the neoadjuvant and adjuvant setting. We have seen the overall survival data, so what are your thoughts? What do you think about this data?

ASR: Well, what an intriguing trial to do, first of all, to compare a question that, as a dose-dense MVAC lover, I have always had – which is better, dose-dense MVAC or gemcitabine cisplatin? This trial incorporated neoadjuvant and adjuvant therapy and patients received either gem-cis or that dose-dense MVAC. Now, the challenge is, if you look at the intent to treat analysis, we saw while there were trends towards improvement with the dose-dense MVAC, it didn’t meet statistical significance. But in the subgroup analysis, looking at those who received neoadjuvant chemotherapy, we do see evidence of benefit from neoadjuvant dose-dense MVAC, better than gemcitabine cisplatin with a significant p-value. But, again, not the primary endpoint of the trial.

YL: So should dose-dense MVAC be the standard of care if you use neoadjuvant chemotherapy? What do we think?

ASR: I personally like dose-dense MVAC, I’ve been using it for a very long time. I think of it as alternating toxicity profiles – if you have a patient who is having problems with thrombocytopenia from the gemcitabine, they may actually do better with dose-dense MVAC. Whereas if they have mucositis from the methotrexate doxorubicin, then they might do better from the gemcitabine cisplatin.

YL: In the trial there were six cycles of dose-dense MVAC; in your practice do you use four, even six?

ASR: Well, it was an interesting choice. Six cycles in the neoadjuvant setting…

YL: It’s quite tough.

ASR: It’s kind of tough. We typically have been giving four cycles of neoadjuvant dose-dense MVAC. Other trials have looked at three cycles but I think they were trying to keep a similar time on chemotherapy. But I wonder, should it be time on treatment or number of times we kill the tumour that really is the best way of comparing regimens?

YL: So, for you, both regimens are quite standard, right, in the neoadjuvant setting?

ASR: In the US both are standard, both are used with high frequency. I would argue I see gemcitabine cisplatin used more across the US, although I personally like dose-dense MVAC and it is my standard of choice.

YL: I think we have a similar trend in Europe. I would prefer dose-dense MVAC but a lot of people are using gem-cis because the tolerance is a little bit better.

ASR: It’s really that historic experience from traditional MVAC which was really rather toxic. I don’t know if you ever had the opportunity to give traditional MVAC?

YL: No.

ASR: I did, back in the day, and it was challenging. I really didn’t like MVAC in the traditional method. Dose-dense really does improve the toxicity profile significantly. So I would just encourage people to give it a try and see how they feel about it.

YL: So, a practice-informing trial but not a practice-changing trial, right?

ASR: It’s definitely practice informing. I think those of us who like dose-dense MVAC can point to that p-value in the neoadjuvant group of patients, saying, ‘Well, it was better in that group.’ Those who like gem-cis can also say, ‘Well, by intent to treat, by the primary analysis, it failed to meet the clinical endpoint.’ So we have evidence for both. I don’t think gem-cis will be supplanted as the standard but with the future of neoadjuvant treatment and combinations, I’d like to see some dose-dense MVAC immunotherapy trials. I don’t know how you feel about that?

YL: Yes, I think again we have a lot of new treatments as well – EV-pembro and a lot of combinations. Most trials use gem-cis in the control arm so that’s why a lot of people now are using cis-gem instead of MVAC but, again, I do agree that dose-dense MVAC is a very good regimen for these patients. So a lot of data this year at ASCO for erdafitinib. Maybe later this year in another meeting, so a great year for erdafitinib. So thank you so much for attending this ecancer discussion. Enjoy, and see you later.