The clinical management of bladder cancer has seen little change over the last three decades and there is pressing need to identify more effective treatments for advanced disease. Low clinical use of neoadjuvant therapies stems from historical limitations in the ability to predict patients most likely to respond to combination chemotherapies. This article focuses on recent molecular and genetic studies, highlighting promising clinical trials and retrospective studies, and discusses emerging trials that use predictive biomarkers to match patients with therapies to which they are most likely to respond. The implementation of predictive genomic and molecular biomarkers will revolutionize urologic oncology and the clinical management of bladder cancer.
Key points
- •
Predictive biomarkers that can identify patients most likely to respond to a given therapy will be of critical importance in advancing bladder cancer management.
- •
Genome-wide DNA and RNA sequencing efforts have revealed bladder cancer to be a heterogeneous disease that harbors alterations conferring sensitivity to targeted agents commonly used in other cancer types.
- •
Bladder cancers have recently been described as having shared properties with several other tumor types, such as lung and breast, and are composed of molecularly distinct subtypes that might predict therapeutic sensitivity.
- •
Immunotherapies targeting the PD-1/PD-L1 axis along with CTLA-4, have shown promising results in recent early phase clinical trials.
- •
Emerging clinical trials that use molecularly guided therapy selection will determine the clinical efficacy of using predictive biomarkers to guide therapeutic decision-making.
Introduction and aims
It is estimated that 74,000 people in the United States and roughly 386,000 people globally will be newly diagnosed with bladder cancer in 2015. Bladder cancer represents a significant global health problem, not only because of its high frequency of occurrence, but also because of high rates of recurrence and the need for routine monitoring via transurethral cystoscopy, which result in significant economic impact. By far the most common type of bladder cancer is urothelial (transitional) cell carcinoma (UCC), which accounts for greater than 90% of all bladder cancer cases. The advent of genomic technologies has led to major advances in molecular testing in cancer medicine for many cancer types with demonstrated clinical benefit in many cases. However, bladder cancer has yet to significantly benefit from these new technologies and its clinical management has changed minimally in the last 30 years.
This article discusses predictive UCC biomarker discovery efforts, and the problems that have impeded their application toward clinical use, while speculating regarding possible future directions that may allow for improved treatment. We emphasize key original articles (and some reviews) and current views, rather than being all-inclusive and thus apologize to authors whose work was not cited. The article discusses the following topics: (1) the clinical need for predictive biomarkers, (2) emerging opportunities for personalized bladder cancer therapy, (3) individual biomarkers of response, (4) new approaches to the classification of bladder cancer, (5) immunotherapy, and (6) clinical trials using molecularly guided therapy selection.
Introduction and aims
It is estimated that 74,000 people in the United States and roughly 386,000 people globally will be newly diagnosed with bladder cancer in 2015. Bladder cancer represents a significant global health problem, not only because of its high frequency of occurrence, but also because of high rates of recurrence and the need for routine monitoring via transurethral cystoscopy, which result in significant economic impact. By far the most common type of bladder cancer is urothelial (transitional) cell carcinoma (UCC), which accounts for greater than 90% of all bladder cancer cases. The advent of genomic technologies has led to major advances in molecular testing in cancer medicine for many cancer types with demonstrated clinical benefit in many cases. However, bladder cancer has yet to significantly benefit from these new technologies and its clinical management has changed minimally in the last 30 years.
This article discusses predictive UCC biomarker discovery efforts, and the problems that have impeded their application toward clinical use, while speculating regarding possible future directions that may allow for improved treatment. We emphasize key original articles (and some reviews) and current views, rather than being all-inclusive and thus apologize to authors whose work was not cited. The article discusses the following topics: (1) the clinical need for predictive biomarkers, (2) emerging opportunities for personalized bladder cancer therapy, (3) individual biomarkers of response, (4) new approaches to the classification of bladder cancer, (5) immunotherapy, and (6) clinical trials using molecularly guided therapy selection.
The clinical need and promise of predictive biomarkers
Standard clinical management for nonmuscle invasive bladder cancers consists of transurethral resection of the bladder, with bacille Calmette-Guérin (BCG) immunotherapy being used in cases with a high risk of progression. For muscle-invasive disease, the current standard of care is radical cystectomy along with lymphadenectomy or radiotherapy. Platinum-based chemotherapy is often recommended, because gemcitabine and cisplatin combination therapy has shown response rates of approximately 38% and 50% in the neoadjuvant and metastatic settings, respectively. These responses are not durable, however, because the overall 5-year survival benefit associated with this neoadjuvant cisplatin-based therapy is a modest 5%. In the adjuvant setting, this survival benefit may be as high as 25%, when compared with patients receiving surgery alone, although these values have been subject to debate. These survival benefits are viewed by many as modest and have led to low use of neoadjuvant treatments clinically. The ability to reliably predict response to platinum-based therapies and other therapies would be incredibly beneficial and would likely result in a change of this current paradigm, increasing the number of patients that are most likely to respond to a given treatment while sparing most unnecessary toxicity.
Predictive biomarkers are molecular or other tumor characteristics that can predict the likelihood of an individual’s response to a given therapy. Several large-scale studies published within the last few years have revealed that bladder cancer is a significantly heterogeneous disease in terms of its genetic drivers, RNA expression profiles, and chemoresponsiveness. The use of genetic profiling has historically been limited to small gene panels and costly molecular diagnostics; however, with the increasing incorporation of next-generation sequencing and other high-throughput technologies in molecular diagnostic laboratories, physicians increasingly have the ability to obtain a more comprehensive understanding of the molecular alterations driving an individual patient’s disease. These molecular characterization techniques have long been used in other cancers, such as breast, lung, and melanoma, to guide therapeutic selection. However, when targeted agents have been trialed in bladder cancer the results have been mixed.
This has resulted in few Food and Drug Administration–approved targeted agents for bladder cancer treatment. Part of the issue with the use of targeted or personalized approaches to bladder cancer treatment is that few clinical trials have enrolled patients based on genomic or RNA expression based biomarkers. A recent review found that of 96 drug-based clinical trials for urothelial carcinoma from January 2012 to January 2015, only 37 (39%) included targeted agents, and of these only 11 (12%) sought to enroll patients based on the appropriate matched molecular or genomic biomarkers. These results highlight the need for increased use of predictive biomarkers in the design of future clinical trials in bladder cancer.
The following sections discuss recent studies that have revealed that most bladder cancers harbor potentially actionable mutations that are likely to respond to existing targeted therapies and molecular profiles that can predict not only untreated patient outcome (prognostic), but also an individual’s responsiveness to specific therapy.
Emerging opportunities for personalized therapeutic regimens
In recent years, several large-scale studies have dramatically expanded on the understanding of the molecular and biochemical underpinnings of bladder cancer. The Cancer Genome Atlas (TCGA) project performed integrative analyses on 131 bladder cancer specimens including whole-exome and whole-genome sequencing, mRNA and miRNA sequencing, and total and phosphorylated protein expression studies. This study, in combination with several others, has provided a more comprehensive picture of the complex molecular landscape underlying bladder cancer development and progression. Perhaps the most important and exciting clinical implication of TCGA data is that it is rapidly being used to redefine how bladder cancers are classified. These new classifications hold tremendous promise to revolutionize the way bladder cancers are treated and how to better predict which patients will respond to various therapeutic options. Newly identified biomarkers will be integral in providing clinicians with the information needed to significantly expand their therapeutic armamentarium for the first time in more than 30 years. TCGA identified potentially actionable alterations in 69% of tumors analyzed. Of these, alterations in the PI3K-Akt-mTOR pathway were seen in 42% of cases and 45% of cases had an alteration in RTK-MAPK pathways.
The complex and heterogeneous array of alterations underlying bladder cancer makes it all the more critical that molecular screening techniques be used in bladder cancer diagnostics. There are already several promising examples of the clinical utility of genomic biomarkers in the treatment of patients with bladder cancer. Genomic biomarker-based clinical trials and promising retrospective studies are discussed next.
Individual biomarkers of response
DNA Repair Pathway Alterations: ERCC1 and ERCC2
Platinum-based therapies are the current mainstay of bladder cancer care, with a subset of patients having remarkable responses. Platinum-based chemotherapies function by forming adducts to DNA and introducing crosslinks. These alterations result in inhibition of DNA replication, leading to cell cycle arrest and the induction of apoptosis. ERCC1 and ERCC2 are members of the nucleotide excision repair (NER) family of proteins, which function to repair DNA damage in the cell. It is therefore not surprising that cancers with high levels of expression of NER genes have been shown to be more resistant to these platinum-based therapies. Conversely, low expression of ERCC1 and ERCC2 has been correlated with responsiveness to cisplatin in bladder cancer. Recently, another study performed whole-exome sequencing on 50 patients before receiving neoadjuvant cisplatin and identified a strong association between responders and ERCC2 mutations. The study authors further show that these mutations could result in increased cisplatin sensitivity in vitro , whereas overexpression of wild-type ERCC2 resulted in increased therapeutic resistance.
TP53
The transcription factor p53 has long been known to play an important role in DNA repair and other cellular processes including the promotion of apoptosis and cell cycle regulation. TCGA described the inactivation of functional TP53 in 76% of samples through a constellation of mutations in TP53 itself, combined with amplifications and overexpression of MDM2. TP53 currently remains an elusive drug target, but there are ongoing clinical trials examining the use of Wee-1 inhibitors, which are thought to sensitize chemoresistant tumors to platinum-based therapies ( NCT01827384 ). That being said, there have been several studies suggesting that p53 expression and mutational status may be predictive of therapeutic response. These studies have produced somewhat confounding results, however, with different studies showing that p53 mutations can confer chemosensitivity or chemoresistance depending on the specific alteration. In the context of bladder cancer, there has been some difficulty defining the role of p53 in therapeutic sensitivity. One retrospective analysis, using immunohistochemistry, showed that patients who had elevated p53 expression and received adjuvant cisplatin saw a survival benefit. Conversely, in a phase III trial that sought to explore the use of p53 expression as a predictive biomarker (again determined by immunohistochemistry) no significant association was shown between p53 expression and methotrexate, vinblastine, doxorubicin, and cisplatin sensitivity. Although TP53 seems to be a major contributor to the development of bladder cancer, more investigation is needed to determine its clinical use as a predictive marker.
PI3-kinase pathway
PI3-kinase pathway alterations were observed in 42% of samples analyzed by TCGA. The alterations seen in this pathway include PIK3CA mutations (17%), TSC1 or TSC2 alterations (9%), and overexpression of AKT1 (9%). Other studies have described cases where TSC1 mutations may confer exceptional sensitivity to targeted therapy, describing the first bladder cancer case of complete response to treatment with everolimus, an MTOR inhibitor. Another recent study described exceptional response in a patient with advanced metastatic bladder cancer to treatment with everolimus and the multitarget receptor tyrosine kinase inhibitor pazopanib. These papers suggest that perhaps activating mutations within the PI3K pathway could serve as biomarkers of response to targeted agents already in use in other cancers.
Receptor Tyrosine Kinases: FGFR3 and ERBB2
Mutations of FGFR3, a receptor tyrosine kinase, have been well characterized in noninvasive and invasive bladder cancers, with approximately 12% of advanced bladder cancers harboring a mutation in this gene. This provides an opportunity to select patients with bladder cancer based on FGFR3 mutation status for potential use of one of the many agents designed to target this gene. One such trial, where patients were screened for FGFR3 mutation status, found that treatment with the pan-FGFR inhibitor BGJ398 saw an exceptional response in a subset of patients with bladder cancer, with four out of five responding to treatment. Tumors in these patients were reduced anywhere from 27% to 48%. Another phase I trial recently found a patient with metastatic bladder cancer harboring an FGFR3-TACC3 translocation showed partial response to treatment with JNJ-42756493, another pan-FGFR targeted agent that has shown promising results in patient-derived explant models harboring various FGFR alterations. It is worth noting that these fusions have been identified as actionable targets across other cancers and in bladder cancer, with TCGA identifying recurrent FGFR3-TACC3 translocations in 3 out of 114 tumors analyzed. Although these studies are currently being expanded, the prospect of using FGFR3 as a biomarker of therapeutic response is promsing.
HER2 is another receptor tyrosine kinase that has been used as a predictive biomarker of response to targeted agents and conventional chemotherapies in urothelial cancers and other cancer types. The role of HER2 in the promotion of bladder cancer has also been explored in several studies and has been associated with increased sensitivity to chemotherapy. Recently, it has been shown that ERBB2 mutations are associated with pathologic complete response (P0) following treatment with platinum-based therapies. This study suggests that ERBB2 could serve as a good genomic biomarker and could provide a basis for selecting patients. Preclinical studies and phase I trials have shown high response rates to HER2-targeted therapies and there are currently phase II trials underway for trastuzumab ( NCT01828736 ) and lapatinib ( NCT00949455 ) in the treatment of bladder cancer.
New approaches to the classification of bladder cancer
Molecular analysis has clearly shown that cancers of specific histologic types are rarely genomically monolithic; infrequently are they defined by a single characteristic mutation or universally predictable in terms of therapeutic sensitivity based on a single genotypic change. Although individual biomarkers may predict response in a single patient or small subset of patients, no single biomarker has the ability to predict every individual’s responsiveness to a given therapy. Patients typically have many co-occurring alterations that can modify their sensitivity to a given compound and may benefit from personalized combination therapies. To better predict the effects these co-occurring events might have on therapeutic sensitivity, several recent studies have aimed to classify tumors across cancer types, describing novel “pan-cancer” subtypes based on shared genetic, molecular, and biochemical features. One of these studies has uncovered remarkable similarities across many cancer types, with bladder cancer standing out as a uniquely heterogeneous and divergently clustered disease type. In this study they evaluated 3527 samples across 12 cancer types, performing integrative analyses across five genome-wide platforms including whole-exome sequencing, DNA copy number analysis, methylation profiling, mRNA sequencing, and microRNA sequencing. Furthermore, data from reverse phase protein array provided proteomic characterization of 131 proteins. Interestingly, bladder cancer primarily clustered into three distinct pan-cancer subtypes, the most divergent classification of any of the cancer types analyzed. Of the 12 cancer types analyzed, five clustered into groups corresponding to their tissues of origin, whereas seven had features that could be clustered into pan-cancer subtypes based on shared molecular characteristics with other cancer types. One of these pan-cancer groups included a subset of bladder tumors, along with samples from the squamous lung and head and neck cancer cohorts. The authors noted that this squamous-like subtype was characterized by p53 mutations, along with amplifications of p63 and enrichment of immune and proliferation pathway features. Aside from the squamous-like subtype, most other bladder tumors clustered into either a group comprised of mostly lung adenocarcinomas, or another bladder cancer–specific subtype, which was comprised primarily of tumors originating in the bladder. These results, suggest that bladder cancer might be looked at through the lens of other cancer types, such as lung, where there is a wealth of data focused on predicting therapeutic response. These sorts of analyses promise to inspire a new wave of clinical trials in which therapeutic decisions are made based on molecular classification rather than traditional pathologic/histologic classification.
Another series of papers has recently been published describing intrinsic subtypes of bladder cancer based on unsupervised clustering derived from genome-wide RNA expression profiling data. These independent analyses of the bladder cancer genome and transcriptome have resulted in the identification of several subtypes that share common expression profiles. The concept of intrinsic subtypes based on unsupervised clustering, has been previously established in breast cancer. This concept of classifying breast tumors based on their molecular taxonomy has been reproduced by many groups independently and is commonly used clinically to inform prognosis and predict response to therapy. Other groups have recently described molecular classification schemes in breast cancer based on normal cell types, and found these may better predict response to therapy than the existing classifications based on tumor-derived profiles. This perhaps suggests that a similar taxonomic approach might be clinically informative in the context of bladder cancer. The classification of these intrinsic subtypes in bladder cancer may have remarkable implications on how patients with bladder cancer are treated and may improve the ability to predict responsiveness to various therapies. A comprehensive description of these subtypes and their role in predicting response in bladder cancer has recently been published.
Immunotherapy: Promising New Horizons
A promising approach that has for years been associated with sometimes remarkable and durable response is cancer immunotherapy. It is hypothesized that the immune system plays a natural role in the prevention of cancer. In addition to being a primary means for combatting foreign pathogens, the immune system exists as a means of surveillance for aberrant processes within one’s own cells. In an actively functioning immune system, when a cell becomes malignant, it displays a variety of metabolic and paracrine cell surface abnormalities that are recognized as abnormal by the innate and adaptive immune systems, causing the cell to be eliminated. Cancer represents a fundamental failure of the immune system to fully execute on its duties as the sentinel to protect against malignant cellular processes, and can derive from either cancer-mediated depression of the natural immune response, or an inherent failure to recognize the cancer cells as needing to be eliminated, because fundamentally they do derive from self, and bear many similar characteristics to one’s own somatic cells. The concept behind cancer immunotherapy is to unleash the powerful cell-regulating potential of this system to effectively target abnormal cells within the body. The concept of immunotherapy has been attempted for many years, but with a few exceptions, until recently, there had been very few breakthroughs.
One of the immunotherapy breakthroughs before activated T-cell therapy and immune checkpoint inhibitors was the use of BCG, a bovine-derived vaccine for tuberculosis, which is injected intravesically into the bladder. The notion of stimulating immune response via the introduction of potent antigens to introduce collateral damage of tumor cells, while conceptually primitive, has a long history first starting in the late 1800s when Coley’s toxin (derived from Streptococcus pyogenes ) was injected intratumorally. Because the body mounts a significant innate immune response to the presence of BCG, the idea behind introducing this into the bladder was to essentially target UCC cells through a sort of bystander effect, by triggering inflammatory processes caused by the presence of BCG (the precise mechanism is not exactly understood, but BCG administration has been shown to activate the innate and adaptive immune systems, and equally targets healthy somatic bladder in addition to cancer cells). Nevertheless, this approach has had significant success in treating UCC, and has been shown to be superior or equivalent to any single chemotherapeutic agent tested to date in terms of reducing progression and recurrence. Associated toxicities are often manageable, because the response is mainly confined to the bladder, given the method of delivery.
However, in the present age of targeted therapy, it should be possible to elicit an immune response in a more focused and specific manner, with minimal collateral damage. The most appealing aspect of targeted immunotherapy is the potential promise for a durable immune response via adaptive immune conditioning, which should prevent recurrence of malignant cells expressing the same antigenic profile. Within the last several years, there have been several breakthroughs in this area, the most successful of which use an armed cytotoxic T-lymphocyte (CTL) response. The characteristics of an effective antitumor immune response involve (1) a mechanism for cancer antigen release, uptake, and presentation by dendritic and other antigen-presenting cells; (2) the recognition of the antigens presented on these cells by appropriate T-cell clones, to prime them for clonal expansion; and (3) the creation of durable immune reserves, in the form of circulating CTLs, which recognize and destroy cells expressing markers of nonself.
One of the most exciting advances in this area has been the development of immune checkpoint inhibitors for programmed cell death protein 1 (PD-1) and its ligand PD-L1 and CTL antigen-4 (CTLA-4). PD-1 is a receptor found on CTLs and some other immune cells that interacts with two ligands, PD-L1 and PD-L2, which when triggered turns off the activated T-cell response and halts the production and release of cytokines. In a healthy immune system, these signaling mechanisms exist to reestablish tissue homeostasis following successful defeat of a foreign pathogenic infection; however, with cancer, tumor cells have evolved to also engage this receptor via synthetic creation of their own PD-L1, which effectively allows them to avoid the wrath of the CTL response.
Several drugs have entered clinical trials to target PD-L1. One of these, MPDL3280A, an anti-PD-L1 antibody, recently received breakthrough therapy status by the Food and Drug Administration because of a 43% objective response rate in patients with PD-L1 + tumors in phase I trials for metastatic UCC. Although less than half of patients saw this clinical benefit, the most exciting aspect of this trial was that even following cessation of treatment after the allotted period, the response seen in patients seemed to be sustained, and through at least the end of the data collection period, a median duration of response was not reached. Phase II trials for this drug have recently finished enrollment. Other exciting antibody-based trials to target the PD-1/PD-L1 axis are ongoing with drugs pembrolizumab, nivolumab (already approved for metastatic melanoma and squamous non–small cell lung cancer), MSB0010718C, and MEDI-4736, among others.
CTLA-4 is a biomarker-based target that functions much in the same vein as PD-1/PD-L1, inhibiting activated T-cell response, via an alternative mechanism through regulation of the T-helper cell signaling that allows for the priming and expansion that ordinarily occurs to build up an army of antigen-specific reactive CTL clones. Preliminary results of ipilimumab, a monoclonal antibody against CTLA-4, in bladder cancer cohorts have been promising, and shown to upregulate immune response in a small cohort of patients with UCC when administered preoperatively.
The other highly promising area of bladder cancer immunotherapy comes in the form of adoptive T-cell transfer (ACT), a proven approach in many tumor types. ACT relies on the extraction and genetic engineering of a patient’s own CTLs to become reactive to tumor-based antigens, through ex vivo priming, clonal expansion, and reinfusion. These enriched CTLs can take the form of traditional α/β T-cell receptors, or be further modified to contain an extracellular domain designed as a mimetic of a tumor-specific antibody, linked to the intracellular domain of the T-cell receptor and costimulatory receptors, to generate enhanced response. This latter approach is known as chimeric antigen receptor therapy. ACT is still in the fairly early stages of testing in bladder cancer, with several ongoing trials, and a new study out of Sweden has just reported a couple of highly promising results from a small cohort of individuals with metastatic UCC, including one patient who demonstrated a complete response. More research into optimizing and expanding this technology is needed to assess its potential in patients with UCC, and although it is a high-cost technology, because of the highly personalized nature of taking each individual patient’s own T cells for the ex vivo modifications, that is also a large part of what makes it so appealing, at least in concept, in the current world of modern precision medicine.
Related posts:
Biomarkers in Urologic Cancer
Urinary Biomarkers for Prostate Cancer
Current Status of Urinary Biomarkers for Detection and Surveillance of Bladder Cancer
Emerging Bladder Cancer Biomarkers and Targets of Therapy
MicroRNAs in Testicular Cancer Diagnosis and Prognosis
The Emerging Role and Promise of Biomarkers in Penile Cancer
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
