Expert Interview of Dr Real: Molecular Pathology and Biology of Urothelial Cancer

Francisco X. Real, MD, PhD
National Cancer Research Center
Madrid, Spain

The therapeutic landscape for urothelial cancer (UC) has undergone significant transformation in recent years, with the addition of PD-1/PD-L1 targeting immune checkpoint inhibitors (ICI) and novel targeted therapies and antibody-drug conjugates (ADCs) adding to the complexity of treatment options available for patients throughout the course of their disease. The optimal sequencing of treatments depends on an understanding of the relevance of different pathological subtypes of the disease and of the latest clinical evidence and guidelines that support the use of different therapies.

By this expert interview, you will receive a detailed grounding of the disease process and management from diagnosis through to salvage treatments of advanced disease. The course will enable clinicians who are involved in the management of patients with urothelial cancer to integrate the latest advances and use of novel therapies for advanced urothelial cancer into safe and effective patient care. 

Transcript

0:16

How does a patient progress from non-muscle-invasive bladder cancer (NMIBC) to muscle-invasive bladder cancer (MIBC) from a pathological perspective? Well, muscle invasion is a relatively simple phenomenon. Tumor cells grow as they go deeper into the bladder wall, or they grow towards the lumen. This corresponds to two different types of growth patterns that are called papillary or solid, and the invasion towards the muscle layer for tumors that are either papillary or solid has a different likelihood of occurrence. For papillary tumors, there's a lesser frequency or tendency to invade muscle. But for solid tumors, the tendency is greater. Invasion into the muscle occurs through not very well-established mechanisms. It can be very readily recognized from a pathological standpoint, but actually, the concept is that the muscle represents a barrier for tumor progression. And we think of that because, in the absence of muscle invasion, it is rare to find lymph node metastases or to find distant metastases. However, there is really a dearth of understanding about how the tumor cells and the muscle interact and communicate with each other. Liquid biopsy is a promising, noninvasive alternative to tissue biopsy for early cancer detection, monitoring, and treatment guidance. Could you please discuss the clinical value, advancements and challenges of urinary tumor DNA (utDNA) and circulating tumor DNA (ctDNA)? This is an absolute revolution. I think that liquid biopsy is going to dramatically change the way patients with bladder cancer are handled. I'm going to refer mainly to muscle-invasive tumors because that's where we have more evidence as of now. But I think it may also impact the management of non-muscle-invasive tumors. In most solid tumors, the work that has been done on liquid biopsy is restricted to plasma DNA. I'm going to refer essentially to plasma cell-free DNA (cfDNA). There are other types of readouts that one can use in liquid biopsy. But referring to tumor cell-free DNA, plasma has been the mainstay in the case of bladder cancer. And actually it may apply to other tumors, but it has been less studied. It is obvious that the urinary output of DNA might be something to explore, and I think that while there is less evidence for this, we're going to see in the next 2 or 3 years a plethora of studies that are going to use urinary DNA. We don't know yet very well how urinary cell-free DNA correlates with plasma cell-free DNA. Regarding the latter, plasma cell-free DNA, I think we're already seeing that it is going to play a fundamental role in the management of patients, both in terms of predicting or telling us about patients who are developing progressive disease after cystectomy, for example. There are good data for that in the context of neoadjuvant chemotherapy, and I think it may actually serve as a guide to decide on perioperative treatment. And the most recent publication in the New England Journal of Medicine, which came out just a couple of days ago, is an example of that. I think the potential is tremendous, and it has tremendous impact on the management of patients from the standpoint potentially of bladder preservation, which is one of the major points that impacts the quality of life of patients with bladder cancer. What is the molecular classification of muscle-invasive bladder cancer (MIBC) and what are its current clinical applications? Muscle-invasive bladder cancer (MIBC) is quite a heterogeneous disease, as is bladder cancer in general. We can classify tumors according to the DNA mutations or genetic alterations that they present, or according to their phenotype. Pathologists have classically done so according to the phenotype, what they see under the microscope. I would say that from a genomic and DNA-based standpoint, two of the critical genomic or genetic factors that impact the classification of bladder cancer and muscle-invasive bladder cancer (MIBC) are whether the tumor goes through the <i>FGFR3</i> <i>mutation</i> pathway, which is mainly the papillary pathway, or whether the tumor goes through an FGFR3 wild-type pathway, which tends to be associated with solid disease. Obviously, <i>p53 mutations</i> we know are very common in solid tumors and are associated with muscle invasiveness. And while not independent, they are a marker of poor prognosis. From a transcriptomic standpoint, the consensus classification has determined the existence of six subtypes of muscle-invasive bladder cancer (MIBC), and these subtypes are distinct. They are well characterized and reproducible. There are tools that allow anybody who runs an RNASeq transcriptomic analysis to classify the tumors. We know that there are three luminal subtypes: papillary, non-specified, and unstable. We have the stromal-rich subtype, which is kind of a mixed bag, and is characterized by the presence of a lot of stroma. But it actually doesn't tell us about the tumor cells; it tells us about the stroma. We have the basal tumors, which we know are associated with a poor prognosis and are more frequent in women. And then we have the neuroendocrine tumors, which do not always match the transcriptomic type with the pathological subtyping. But we know that they are associated with a very poor prognosis. But of course, the important thing is not only prognosis, it's whether these classifiers allow changes in management or treatment decisions. And at this stage, I cannot say that there is a direct impact of tumor classification on treatment decisions. What are the most valuable biomarkers in advanced or metastatic urothelial carcinoma, and what standardized methods are currently used for the detection? Honestly, I think that the biomarkers that we have now for bladder cancer are still relatively poorly developed in comparison with other solid tumors.<i>FGFR3</i> <i>mutations</i> definitely guide therapy for FGFR3 inhibitors, and there is a good basis to say that in the absence of mutations or fusions (which, as you know, are relatively rare but do occur in 3 to 5% of patients), the likelihood of response, in the absence of these genetic alterations, to FGFR3 inhibitors is really nil. Regarding the other treatment decisions in muscle-invasive bladder cancer (MIBC), I think that, I have to say that, the markers that we have are relatively poor. Even the immune checkpoint inhibitor-associated biomarkers, such as <i>PD-L1</i>, we know are not fantastic or as good as in lung cancer. So there is still a lot of development to be done in this area, I think, before they can impact the management of patients. It appears that certain biomarkers exhibit differential expression between early-stage and late-stage urothelial carcinomas, particularly <i>FGFR</i> and <i>HER2</i>, which are relevant to biomarker-driven precision therapies. Could you please provide further insights into this observation? Yeah, the best evidence is actually for <i>FGFR3</i>. <i>FGFR3</i> is an oncogene. We have hotspot mutations. The mutations are activating of the kinase. Interestingly and paradoxically, the mutations are associated mainly with non-muscle-invasive tumors, actually among non-muscle-invasive tumors with the low grade, the ones that are less aggressive. So what this tells us is that it identified <i>FGFR3</i> <i>mutations</i>, and identified a pathway of tumor evolution, which in itself is insufficient to determine muscle progression. The accumulation of additional alterations, such as <i>p16</i> <i>deletions</i> or <i>p53</i> <i>mutations</i>, is what actually allows these <i>FGFR3</i>-driven tumors to progress to muscle invasion. So we can say that <i>FGFR3</i> <i>mutations</i> are a driver for non-muscle invasive tumors, whereas other drivers are required for these tumors to progress to muscle invasion. Regarding <i>HER2</i>, the data are less clear. I would say that <i>HER2</i> is clearly amplified more frequently in muscle-invasive tumors than in non-muscle-invasive tumors, but where <i>HER2</i> plays a role is still relatively unclear, as it is, I believe, still relatively unclear. What is the differential impact of overexpression at the protein level versus overexpression at the mRNA level versus amplification? For some tumors, like gastric cancer or breast cancer, we have been able to dissect very carefully. What is the clinical significance as a biomarker of these different readouts? I think for <i>HER2</i>, we still have some way to go in bladder cancer. Of course, the development and again the recent ESMO-associated publication in the New England Journal of Medicine, of the trial with antibody drug conjugates (ADCs) targeting <i>HER2</i> is definitely going to change the landscape of these therapies, and I think it's going to accelerate the use of this protein or this gene as a biomarker. The use of immunotherapy and antibody drug conjugates (ADCs) has been approved for multiple indications in the treatment of urothelial cancer (UC). Are there any predictive biomarkers available to identify potential responders who would benefit from the novel therapies? The simple answer to that is no. I don't think that we have yet reliable biomarkers. There are a few retrospective studies. Let me go back one step. The field of biomarkers is full of retrospective studies, which are not necessarily highly reproducible due to all the bias associated with retrospective studies, selective biases related to sample acquisition, to clinical availability of data, to problems with detecting the biomarkers in all samples, etc. So I think that we still have a long way to go in this regard. And I think that while it is important to identify these biomarkers for <i>Nectin4</i> or <i>Trop2</i>, the data is still relatively scarce. It hasn't been validated across assays in different institutions, and the data for <i>HER2</i>, for example, is still too early, I think, to say that there is sufficient data to use them as biomarkers in clinical management of patients.