Andrea, you were here at the recent IBCG retreat in Houston. We talked about biomarkers, we talked about how we can actually use these biomarkers in the systemic setting, in the metastatic urethral cancer. Of course you did a lot of research with your team on a background towards that, so really looking forward to hearing and sharing with the worldwide audience what we discussed and what you discussed, especially at the retreat. So, take it away.
Andrea Necchi: Thank you, Ashish. Thank you and thank UroToday for inviting me. It is always a pleasure to share ideas and to share the latest developments in bladder cancer with you and to discuss the future application of these developments. As you said, Ashish, we discussed quite a lot around the issues of biomarkers, and the biomarker was the primary focus on this year's IBCG meeting. I will navigate very briefly through the biomarkers that are, today, ready for prime time for use for orienting immunotherapy choice or targeted therapy choice or using biomarker as a way of monitoring response to treatment or anticipating disease relapse after a radical surgery or radical treatment. Let's start with immunotherapy biomarkers. I will tell you something that is not actually quite new because the immunotherapy related biomarkers are still referring to tools like the TMB, the tumor mutational burden, which is a pretty well-known biomarker in solid tumor oncology and is often considered as a surrogate of tumor neoantigen load, as the way of anticipating or predicting a possible response to immune checkpoint inhibitors, either use a single agent or combination therapy.
You see here on the left-hand side, in this slide, that bladder cancer stands in the middle with regards to TMB values, pretending a possible intermediate to good response as an average inpatient receiving immune checkpoint inhibitors. Most importantly, what I found very, very important, interesting with regards to the possible application of TMB is that we are dealing with a continuous value. If you see the slide on the left-hand side here, I circled the most important point related to urothelial cancer, this slide is relative to the court of foundation medicine, and when looking at the different solid tumor types, you see that there is a steady increase in the TMB values, so with regards to the activity and to the efficacy of immune checkpoint inhibitors. So, there is the proportional association between increasing TMB values and efficacy of immune checkpoint inhibitor. So this is a good point to know.
In particular, when applying or when using or envisioning in the future or envisioning in a routine practice a possible use of pembro or any other checkpoint inhibitor, so inpatient presenting with a very high level of TMB as it may happen, like 90 mutations per megabyte or 80 mutations per megabyte, these values are clearly pertaining response to checkpoint inhibitors. As stated here in the slide, we have a lot of uncertainties because there is still a jungle with regards to the way we measure TMB, the tools that are used today in routine practice or that are available or reimbursed depending on the various geographical areas for patients with bladder cancer as well as for the other cancer, and the heterogeneity, tumor heterogeneity, temporal heterogeneity, and heterogeneity between the two.
So we are still at the very beginning when translating these notions to the real world. For sure, what we aim to is something like this. Recent publication on New England Journal of Medicine with MSI-high mismatch repair deficient cancer receiving checkpoint inhibitors neoadjuvantly. You see here that there is clearly almost 100% universal or 100% opportunity of response to this tumor portending the possibility of saving or sparing major surgery in most of these tumor types, including a few cases with urothelial cancer. But we are quite far for most of the patient diagnosed with bladder cancer from this reality. And we are still to face with problems related to other biomarkers like PD-L1 expression with regards to immunohistochemical assessment of PD-L1 expression in urothelial cancer. You see that here, the performance of the various antibodies is pretty much similar, with the only exception of SP142, which is the Roche antibody related to atezolizumab, which stands a little bit outside of this performance.
But in general, we are not very much happy when using immunohistochemistry in assessing PD-L1 except for a few indication today in bladder for choosing the right treatment for this patient. Finally, we have the ctDNA. ctDNA is considered now universally considered now the new holy grail of solid tumors. In fact, the data, also including the data on bladder cancer, are pretty much exciting. The clinical application of ctDNA are shown here, are summarized here. In particular, what is now pretty well established in bladder cancer is the possibility of predicting the prognosis of the patients after a diagnosis or after a new adjuvant therapy or after a surgery and maybe predicting the response to adjuvant immune checkpoint inhibitors. This slide actually summarizes the latest findings, the latest developments. With the use of ctDNA from the left-hand side, the huge work done by the Danish colleagues, Lars Dyrskjøt and his team, down the road to IMvigor data that are pretty well known and will be updated soon at ESMO meeting.
But anyway, ctDNA as a way of predicting the benefit of adjuvant immune therapy and the optimal patient to select for adjuvant strategy of immune therapy. Much less data and much less results are today available for patients with more advanced disease, patient with metastatic disease. For patient with metastatic disease, the use of ctDNA is still a pretty much concern because there are realities, geographical realities like in the United States where the signature tests or other ctDNA tests are used and are available as compared to many other geographical areas. But the point is that, for metastatic patients, we still have uncertainties regarding the benefit of using or using too frequently the ctDNA assessment while on treatment. This is a case report recently published on the way of predicting the imaging relapse on a patient receiving pembrolizumab second-line therapy. But in general, we are not very much happy today based on the few data that we have today in using too much or in suggesting the use of ctDNA too frequently because the risk of errors in decision making on losing or burning the therapeutic possibilities is still too high.
We have a possibility regarding targeted therapy in bladder and targeted therapy as to do with NGS opportunities, for opportunities for the patient to have their tumor assessed through NGS tests. This is a summary based on the last year, two years ago actually, Milan retreat meeting, where we convened on the fact that the NGS should be considered for all patients with at least a diagnosis of muscle-invasive bladder cancer. This notion and this consensus regarding this statement was actually confirmed at this year's IBCG meeting in Houston. So, the possibility for patient with at least a locally advanced disease to get any kind of assessment, a molecular assessment for possible use of targeted therapy is very important. Why so? Because we already have targeted therapy opportunities, in particular, focusing on FGFR3 genomic alteration. This is a picture of the latest data, clinical data on the box on the right-hand side from the newer pan-FGF receptor inhibitors. Very exciting.
Response rated are in the range of 40% to 50%. Small patients, phase one, two studies are pretty much exciting, and these are all drugs that are coming into the ground of the post-erdafitinib in the world. Erdafitinib was the first pan-FGF receptor inhibitors the FDA and EMA approved for patients whose tumor is harboring any kind of FGF receptor 2 or 3. A genomic alteration based on this study, the THOR study, showing an increase in survival where the erdafitinib is compared to investigator choice chemotherapy in patient who had already received one chemotherapy or platinum chemotherapy and an immunotherapy drug. Much less data, much less excitement was reported for patients who were immune naive with the use of targeted therapy, and there is still uncertainty regarding the immune naive patients where it is the case or not to use targeted therapy.
Other biomarkers for targeted therapy. Let's say, if we consider somehow ADCs as being a targeted therapies, in fact for certain ADCs we do have to deal with the targeted therapy. For enfortumab, we are still dealing with all-comer approach based on this outstanding data shown here for all comers. Nectin-4 is still an uncertain biomarker for benefit of an enfortumab a lot in efficacy. But we do have, as I mentioned, biomarkers assessed by immunohistochemistry. So easy to assess like HER2, for which we do have already a plethora of newer ADCs like Trastuzumab or other ADCs that have been granted accelerated approval as in the case of trastuzumab deruxtecan in all tumors expressing immunohistochemically, HER2 with 2+ or 3+ staining including bladder cancer. So we do have, at least in the United States, unfortunately not in many other geographical areas, we do have a possibility in later-line therapy for considering a biomarker-oriented approach after EV-Pembro, after chemo, after erdafitinib, so for one further efficacious treatment in this patient.
Finally, the monitoring of tumor resistance is pretty much exciting today. MRD is a concept that there was a rise and reported and using the routine practice historically by the hematologist. Minimal residual disease assessment stands for the assessment of cancer that cannot be detected with current imaging tools, with the radiology, in a way of anticipating disease relapse or failure of any kind of treatment that the patient is receiving. So multiple uses may imagine for routine practice, including bladder cancer. In fact, we have a lot of data related to ctDNA. As I mentioned before, on the left-hand side, the meta-analysis, the pan-tumor meta-analysis of the use of ctDNA neoadjuvantly in predicting the pathological response in solid tumors. On the right-hand side, again, the work of Lars Dyrskjøt and colleagues, one of the latest works showing the survival improvement in patients with a clearance of the ctDNA after surgery, after cystectomy in bladder cancer patients or after neoadjuvant chemotherapy and radical cystectomy.
These are the latest data in the perioperative setting from the NIAGARA, presented by Tom Powles at this year's main ASCO meeting, again showing that the clearance of the ctDNA was fairly prognostic. The baseline DNA, ctDNA negativity was prognostic, was associated with the good prognosis for this patient, and the ctDNA clearance post-neoadjuvant therapy or neoadjuvant chemoimmunotherapy and surgery was clearly prognostic. And this is an example of many outside of the standard chemotherapy backbone. One case of patient included in one of our studies internally, the NureCombo study that we published last year in JCO, receiving neoadjuvant nivo-braxine. Again, here, the clearance of the ctDNA while on treatment portended a very good response at the pathological level, at the level allowing the patient to skip the radical cystectomy and start the maintenance immune therapy approach. So exciting opportunities also for bladder sparing.
One of the thing that is exciting to me, it is another way of considering the MRDSS, but not in the case of having a tool that is quantifying a value of the circulating tumor DNA, but as a way of detecting the newer mutations when particular patient receiving targeted therapy, mutation that could be associated with resistant mechanism development to certain targeted therapy, and mutation that can be targeted with further targeted therapy. So, a liquid biopsy is a way of overcoming the limitation of tissue biopsy, and this is the case of the pemigatinib study, another pan-FGF receptor inhibitors. These six cases were assessed sequentially, longitudinally during treatment. These six patients developed resistance to pan-FGF receptor inhibitors and we found the emergence of the resistant mutation in any of these patients, in particular, including gatekeeper mutations that have been reported in the literature, has been associated with erdafitinib or other FGF receptor inhibitor resistance.
Now, the new generation of FGF receptor inhibitors, like those that I've shown you in previous slide, are able to overcome the occurrence of these mutations as a way of overcoming the immune resistant mechanism and are classically related to the first generation FGF receptor inhibitors. For patient with metastatic cancer receiving EV-Pembro, for example, we have initial data from observational studies conducted in the United States, like in this case, in this poster of the ASCO meeting, showing again that a ctDNA, a classically evaluated ctDNA with Signatera was associated also in metastatic patients with the benefit from EV-Pembro treatment. Again, here, another summary of the data with the ctDNA assessment and mostly linked to the Danish work and mostly linked to the way of anticipating the disease relapse in the perioperative setting.
In general, as a summary, we are pretty much excited about the use of ctDNA in the perioperative setting. And then, for sure, there is a way of skipping major surgery in certain patients or let's say integrating the definition of clinical response that we have discussed so many times, Ashish, within IBCG and in many meetings. But the point now is that we need to have more data on metastatic patients and we need to validate prospective studies the validity of these tools. Also, considering the fact that they are quite expensive tools, in general, healthcare system cannot afford at a large scale these tools for all comers. So, we as investigators, as academic investigators are asked to be smarter enough to digest all of these data that they briefly present to you and to apply this data to the patient bed and to the single patient level that we see on a daily practice. This is a challenge for the next generation physician, scientist, and this is also the challenge for the next generation clinical trials that we will use in any clinical setting.
Ashish Kamat: Thanks so much, Andrea. As always, you were able to take such a complicated topic and, in a succinct manner, give the data. Again, this was a real important part of the discussion we had at the recent meeting, and as you showed this collaboration between us, our groups across continents is something that's really driving the field forward, and looking forward to the meeting in Milan in a couple of months. You presented everything so well, but I want to put you on the spot a little bit and ask you briefly if you could share with us your thoughts on all these biomarkers, all these predictive abilities, ctDNA, MRI, et cetera. These are some things that we study on a population level, and we average out based on clinical trials, et cetera. What is your thought process behind using this data that we glean from these retrospective studies, prospective studies, but then using AI-based or machine learning-based algorithms to make it truly personalized for the patient? Where do you think we are now and where are we heading towards personalizing it for the individual patient sitting right in front of you?
Andrea Necchi: That's an outstanding question, Ashish, and I think that you touched very nicely the point of today's research and today's aim to cure the patient. Because I am a bit pessimistic with this regard, so I think that we will never result in identifying a tool with prospective studies, as you said, on population-based studies, on tools that will clearly make a difference, will clearly be transformative in the way of selecting the patient and rising the bar of therapeutic success of almost 100%. This is our aim. The TMB is the typical example. TMB has never emerged from any clinical trial, prospective or retrospective study, post-hoc analysis from trial has been clearly associated with benefit with immune checkpoint inhibitors. But as a matter of fact, when I see on everyday clinic a patient presenting with a TMB or let's say 80 or 90, I'm pretty much sure that the patient is almost 100% chances of responding pretty well to immune checkpoint inhibitor.
As you said, we probably need a newer way of designing clinical trials, and AI may help with this regard by using design and may be pretty well-set and well-shaped in the answering this question, and maybe the overall, the trial design, which is aimed to validate certain biomarkers has to change a little bit because we have to identify the outliers. We have to identify the outliers when dealing with TMB, when dealing with this and the CPS PD-L1 expression, maybe in the future Nectin-4 expression, certain other biomarkers.
So with only a few exceptions like mismatch repair deficiency or so, for all the other biomarkers, we need to access, we need to identify the outliers. And clinical trials, classically to conceived clinical trials are not pretty well-shaped to answer the question of identifying the outliers, so AI will be pretty well shared. Honestly, I'm not aware of currently available robust clinical trials that will support or will consolidate this notion. So, for sure, everything should be proven prospectively. But we have today the tools, the AI tools at least to invest on this to change our mind and to change the way we currently include patient in trials. We currently conceive clinical trials in bladder as well as many other solid tumors.
Ashish Kamat: I agree with you completely. Andrea, always a pleasure. Thank you for taking the time. Looking forward to seeing you soon.
Andrea Necchi: Thank you. Thank you very much.