Novel Biomarkers to Predict Response a nd P ro g n o s i s in Lo c a l i z e d B l a d d e r Ca n c e r Ilaria Lucca, MDa,b, Michela de Martino, PhDa, Tobias Klatte, MDa, Shahrokh F. Shariat, MDa,c,d,* KEYWORDS Nonmuscle invasive bladder cancer Diagnostic biomarkers Prognostic biomarkers DNA RNA Protein Urine
KEY POINTS Urothelial carcinoma of the bladder (UCB) is the most expensive solid tumor to treat because of its high recurrence rate and the need of continued cystoscopic surveillance. Cystoscopy has a high sensitivity and specificity, but is a costly and invasive procedure; urinary cytology is noninvasive and highly specific and shows very low overall sensitivity, in particular in low-grade nonmuscle invasive bladder cancer (NMIBC). Several biomarkers may have better sensitivity than voided urinary cytology, but they are not accurate enough to replace cystoscopy and cytology. NMP22, BTA, ImmunoCyt/uCyt1 and UroVysion/FISH have been approved by the Food and Drug Administration for screening and follow-up of patients with NMIBC in combination with cystoscopy. The long interval required for validation, testing, and approval of the assays and the lack of standardization could explain the present failure of biomarkers to predict NMIBC development, recurrence, and progression.
Urothelial carcinoma of the bladder (UCB) is the most expensive solid tumor to treat because of its high recurrence rate and need of continued surveillance, with an estimated annual cost of $3.7 billion in the United States.1 More than
400,000 patients are newly diagnosed with UCB worldwide every year, and more than 150,000 die from the disease.2 The standard evaluation of a suspected UCB consists of cystoscopy and urinary cytology.3 Although cystoscopy has a high sensitivity and
Funding Sources: Development fund of the CHUV-University Hospital, European Urological Scholarship Programme (I. Lucca); Oesterreichische Nationalbank Anniversary Fund (project number: 15363/2013), Medical-Scientific Fund of the Mayor of the City of Vienna (project number: 14069/2014) (T. Klatte); Nil (M. de Martino, S. Shariat). Conflict of Interest: Nil. a Department of Urology, Comprehensive Cancer Center, Vienna General Hospital, Medical University of Vienna, Wa¨hringer Gu¨rtel 18-20, Vienna A-1090, Austria; b Department of Urology, Centre Hospitalier Universitaire Vaudois, Rue du Bugnon 46, Lausanne CH-1010, Switzerland; c Department of Urology, University of Texas Southwestern Medical Center, 1801 Inwood Road, Dallas, TX 75235, USA; d Department of Urology, Weill Cornell Medical College, New York-Presbyterian Hospital, Cornell University, 1300 York Avenue, New York, NY 10065, USA * Corresponding author. Department of Urology, Comprehensive Cancer Center, Vienna General Hospital, Medical University Vienna, Wa¨hringer Gu¨rtel 18-20, Vienna A-1090, Austria. E-mail address: [email protected] Urol Clin N Am 42 (2015) 225–233 http://dx.doi.org/10.1016/j.ucl.2015.01.004 0094-0143/15/$ – see front matter Ó 2015 Elsevier Inc. All rights reserved.
urologic.theclinics.com
INTRODUCTION
226
Lucca et al specificity, it is a costly and invasive procedure, with a risk of complications, such as urinary infections, pain, and bleeding. Urinary cytology has a low overall sensitivity rate of only 33% to 48%, but a higher specificity of 86% to 90% in the detection of nonmuscle invasive bladder cancer (NMIBC).4,5 Its predictive accuracy improves in high-grade tumors to approximately 90%, whereas it does not go beyond 60% in lowgrade NMIBC.6 Urinary cytology after bladder washing seems to have a higher sensitivity than from voided urine.7 In UCB, the TNM staging together with various prognostic models, such as the European Organization for the Research and Treatment of Cancer (EORTC) and Spanish Urological Club for Oncological Treatment (CUETO) scoring systems, allow defining the risk of recurrence and progression after the surgery.3,8 However, among patients with UCB staged similarly, there is still a considerable variation in rates of recurrence, disease progression, and response to treatment. Therefore, biomarkers are needed to improve prognostication. Intense work is being done in the field of NMIBC biomarkers with the main goals of reducing the number of invasive cystoscopic evaluations through early diagnosis and identifying those patients with a higher risk of disease progression. The identification of valuable NMIBC biomarkers would improve our understanding of the biological pathways involved in the genesis and progression of the tumor, but would also add information about UCB outcomes. A biomarker is an indicator of biological and pathogenic processes or pharmaceutical responses to a therapeutic intervention.9 The ideal
Fig. 1. Classification of biomarkers for NMIBC.
biomarker should be noninvasive, rapid to analyze, objective, easy to perform, sensitive, specific, reproducible, and, above all, cost-effective. However, to date, the available biomarkers are all lacking one or more of these characteristics. Several biomarkers showed a higher sensitivity than voided urinary cytology, but they are not accurate enough to replace it. Indeed, although several studies have been published on NMIBC biomarkers in recent years, none of them is recommended for use in clinical practice. The classification of UCB biomarkers is relatively complex. They could be classified according to (1) the sample used for their identification, (2) their objective, (3) the biomolecule, or (4) the type of measurements applied (Fig. 1).10 Because the literature on this topic is vast, the aim of this review was to provide a summary of the most relevant biomarkers investigated in the last period as predictor of NMIBC, recurrence, and aggressiveness.
DIAGNOSTIC BIOMARKERS FOR EARLY DETECTION OF NONMUSCLE INVASIVE BLADDER CANCER Diagnostic biomarkers aim to detect the disease in symptomatic patients or in those having an increased risk, if possible at early stages.11 NMIBC is an attractive tumor for screening because of its well-known risk factors and its favorable outcomes when diagnosed at early stages. Macrohematuria is the most common symptom, but it is not specific and is not related to the pathologic stage. Madeb and Messing12 screened 1575 men aged 50 years or older without a history of UCB,
Biomarkers for Localized Bladder Cancer hematuria, or previous treatment with pelvic radiation therapy for hematuria. Although 6% to 10% of men with hematuria were diagnosed with UCB, hemoglobin dipsticks had a positive predictive value of only 8.1%. It has been hypothesized that a combination of home hematuria urinary testing and a panel of biomarkers may improve the sensitivity and specificity of the screening. The Bladder Cancer Urine Marker Project (BLU-P) tested a total of 1747 Dutch men with a home hematuria test followed by a molecular marker panel, which included nuclear matrix protein 22 (NMP22), microsatellite analysis (MA), the fibroblast growth factor receptor 3 (FGFR3) mutation snapshot assay, and a custom methylationspecific (MLPA) test.13 The specificity and sensitivity of any positive molecular test were 95.9% and 80.0%, respectively. However, of the 71 men who finally underwent cystoscopy, only 4 UCBs and 1 upper tract urothelial carcinoma were detected, showing a lack of benefit of this combined screening approach in a large unselected asymptomatic population.
Protein Biomarkers Several studies have reported the association of urinary protein biomarkers and UCB, leading to the Food and Drug Administration (FDA) approval of NMP22, bladder tumor antigen (BTA), and ImmunoCyt/uCyt1 for screening and follow-up in combination with cystoscopy. NMP22 is a protein found in the nuclear matrix and is involved in replication and gene expression. The sensitivity and specificity of the urinary NMP22 test to detect UCB is approximately 68% and 79%, respectively.14–16 The test is commercially available and detects urinary levels of the nuclear mitotic apparatus protein, which is elevated in the urine of patients with UCB. However, a high rate of false-positive tests has been reported, often due to macrohematuria, urinary tract infections, and instrumentation. Ponsky and colleagues17 showed that benign inflammatory or infectious conditions, renal or bladder calculi, current or recent history of foreign body in the urinary tract, any bowel interposition segment, and other genitourinary cancers are responsible for reducing the positive predictive value and the specificity. Based on an experimental model, Miyake and colleagues18 suggested that urinary NMP22 assays may diagnose UCB by counting the amount of cell turnover as consequences of surface shedding from bladder tumors rather than identifying a specific tumor antigen. That would explain the false-negative NMP22 tests in those with reduced urinary cellularity due to the characteristic of the
tumor or an error in sampling. Moreover, concentrated urine may be also responsible for falsepositive tests because of an overestimation of NMP22 levels.19 Taken together, NMP22 cannot be considered an alternative to cystoscopy. The BTA test detects the human complement factor H-related protein, which seems to prevent cell lysis in UCB.20 The original BTA test (Bard Diagnostics, Redmond, WA) was taken off the market because of its very low specificity and sensitivity.21 Two tests are currently available, the BTA stat and BTA TRAK (Polymedco, Cortlandt Manor, NY). Although the first is qualitative and gives an immediate result, the second is quantitative and requires experienced laboratory processing. The overall BTA sensitivity is higher than urinary cytology (61%–67%), but the specificity seems to be lower (75%–78%).4,22 The presence of blood in the urine interferes with the BTA tests, causing a high rate of false-positive results due to the cross-reactivity to antigens on red blood cells.23 Some investigators have suggested the use of BTA as serum biomarker, measuring complement factor H introduced by bleeding.24 ImmunoCyt/uCyt1 (Scimedx, Denville, NJ) is approved by the FDA for the follow-up of patients with NMIBC. It combines 3 monoclonal antibodies and operates as a second-level test in case of atypical urinary cytology. Odisho and colleagues25 performed ImmunoCyt/uCyt1 tests on 506 voided urine samples with atypical urinary cytology. For low-grade UCB, the sensitivity, specificity, and negative predictive value were 75%, 50%, and 82%, respectively. For high-grade UCB, the values were comparable (74%, 44%, and 79%, respectively). Moreover, Teˆtu and colleagues26 showed a significant improvement of the urinary cytology sensitivity when combined with ImmunoCyt/ uCyt1, with a sensitivity of 78% in Ta, 83% in T1, and 100% in carcinoma in situ.
Chromosome Biomarkers Another FDA-approved urinary biomarker is UroVysion (Abbott Laboratories, Abbott Park, IL), a multitarget fluorescence in situ hybridization (FISH) assay that detects aneuploidy in chromosomes 3, 7, and 17 as well as focal loss of 9p21 containing the p16 tumor-suppressor gene.27 A meta-analysis of 14 studies involving 2477 UroVysion-FISH tests found an overall sensitivity of 86% (82%–89%) for UroVysion compared with 61% (56%–66%) for urinary cytology.28 However, positive FISH did not reach the level of evidence for the initial diagnosis of UCB. In patients with NMIBC with positive cytology but negative cystoscopy, the prediction of disease occurrence or
227
228
Lucca et al recurrence is challenging. In a study investigating 243 patients with NMIBC with atypical urinary cytology, a positive UroVysion-FISH test was a significant predictor of recurrence (hazard ratio 2.35, P 5 .001) in multivariate analysis.29 Finally, Seideman and colleagues30 showed that patients with NMIBC with atypical urinary cytology and positive UroVysion-FISH had a shorter time to recurrence compared with those with a negative UroVysion-FISH (12.6 vs 17.9 months), even when cystoscopy was negative.
DNA Methylation Biomarkers DNA methylation is one of the most studied epigenetic mechanisms, which commonly occurs in CpG-rich regions that are located in the promoter regions and lead to gene inactivation.31 Because CpG hypermethylation does not occur in normal cells, many methylated genes have been linked to UCB pathways.32–38 The detection of methylation biomarkers has several advantages over conventional genetic markers because of their high degree of specificity, which is important when DNA is available in limited quantities (eg, from biological fluids such as urine). Moreover, they play a role in both diagnosis and prognosis of UCB. Su and colleagues39 built a panel of biomarkers, including a combination of a transcription factor (SOX1), a specific LINE-1 element, and interleukin-1 receptorassociated kinase 3 (IRAK3), which could detect a higher rate of early recurrences than urinary cytology and cystoscopy. DNA methylated biomarkers, such as OTX1, ONECUT2, and OSR1, have been studied in combination with FGFR3 assays, showing a sensitivity of 79% and a specificity of 77% in patients with NMIBC.40 Because DNA methylation is involved in UCB pathogenesis, numerous promoters of DNA hypermethylation are being investigated (ie, tumorsuppressor genes, proto-oncogenes, genes involved in cell adhesion, and genes regulating the cell cycle).41,42 In a recent study, histone deacetylases (HDAC)-1 and HDAC-2 overexpression was associated with the development of high-grade NMIBC, suggesting a potential role of HDAC inhibitors as chemotherapeutics.43 Moreover, downregulated KLF4 has been found to promote DNA hypermethylation and may be associated with a higher risk of early recurrence and disease progression.44
Micro RNA Biomarkers Micro RNAs (miRNAs) are short noncoding RNA molecules of 18 to 22 nucleotides that are able to modulate posttranscriptional protein expression. Upregulation or downregulation of
miRNAs may influence UCB tumorigenesis by acting as oncogenes or tumor-suppressor genes. A total of 2216 miRNAs have been registered so far.45 Most miRNA studies focus on tissue samples, whereas only one-third is on urine samples. Low-grade UCB is characterized by miRNA downregulation, whereas in high-grade UCB, there is typically miRNA upregulation.46 A recent metaanalysis identified 15 miRNAs as the most promising diagnostic and prognostic UCB biomarkers, of which 11 were upregulated (miRNA-96, miRNA-138, miRNA-126, miRNA-182, miRNA143, miRNA-222, miRNA-21, miRNA-133b, miRNA-518c-5p, miRNA-452, and miRNA-129) and 4 were downregulated (miRNA-200c, miRNA-99a, miRNA-100, and miRNA-29c).47 Among them, miRNA-99a expression was inversely correlated with the aggressiveness of UCB, suggesting a role of miRNA-99a as tumor suppressor.48 Also miRNA-100 might act as a tumor suppressor. Indeed, when artificially reintroduced in tumor tissue, it is able to suppress cell proliferation and motility, inducing cell-cycle arrest in vitro.49 Finally, some promising miRNAs, such as miRNA-141 and miRNA-639, have been recently studied in plasma, but final results showed no differences in serum levels between patients with UCB and controls.50
Messenger RNA Biomarkers Survivin, also known as BIRC5 or EPR-1, has a pivotal role in promoting cell proliferation and preventing apoptosis. Its sensitivity and specificity to predict UCB have been estimated at approximately 76% to 78% and 82% to 95%, respectively, and when survivin tests are combined with urinary cytology, the accuracy increases significantly.51–53 Moreover, patients with NMIBC expressing high survivin levels have shown a lower risk of disease recurrence.54 Another interesting mRNA biomarker is Aurora A kinase (AURKA), a cell-cycle–associated serine/threonine kinase involved in genetic stability. The AURKA overexpression seems to be specific of UCB tissue and in voided urine samples may be associated with higher UCB stage and grade, and worse survival.6
PROGNOSTIC BIOMARKERS IN NONMUSCLE INVASIVE BLADDER CANCER In NMIBC, presence of carcinoma in situ, tumor stage and grade, tumor size, and numbers of lesions are established variables to predict disease recurrence and progression. It has been estimated that 10% to 20% of patients progress to a muscle-invasive stage.55,56 New prognostic biomarkers that identify patients with a high risk
Biomarkers for Localized Bladder Cancer of progression, and may therefore be candidates for a more aggressive therapy, are needed.
Inflammatory Biomarkers Inflammation plays a pivotal role in the development and progression of UCB.57,58 It is commonly accepted that the tumor microenvironment triggers the release of cytokines, subsequently influencing the systemic inflammatory response.59 Measurements of systemic inflammatory response represent valuable, reproducible, and inexpensive prognostic biomarkers. According to a recent meta-analysis, inflammatory cells, UCB-costimulatory molecules, serum cytokines, and the C-reactive protein were all significantly associated with survival.60 Some inflammatory measurements have been combined in ratios to improve their prognostic value. A high neutrophil-to-lymphocyte ratio (NLR) has been associated with poor survival and higher risk of disease recurrence in NMIBC. Mano and colleagues61 estimated a 3-year progression-free survival rate of 61% in patients with NMIBC with a preoperative NLR greater than 2.41, compared with a rate of 84% in patients with an NLR of 2.41 or less (P 5 .004).
Cell Cycle and Adhesion Markers In the past decade, several molecular biomarkers have been identified as potential predictors of UCB progression to an invasive stage.62 The p53 protein is a well-known tumor-suppressor protein with a crucial role in the prevention of genomic mutations. Besides its role in predicting early recurrence, several studies showed an association between p53 overexpression and UCB progression.63,64 Shariat and colleagues65 studied p53 in combination with other molecular markers (pRB, p21, and p27). The investigators found that a high number of altered biomarkers was significantly associated with an increased risk of disease recurrence (P 5 .011) and progression (P 5 .005). Deletions of the long arm of chromosome 9 and mutations in FGFR3, H-RAS, and PI3K also may influence the development of UCB.66 Lindgren and colleagues67 found higher frequency of FGFR3 mutations in low-grade NMIBC (80%) than in high-grade NMIBC (less than 10%), suggesting a protection role of FGFR3 against disease progression. On the other hand, others failed to show that deletions of the long arm of chromosome 9 and mutations of HRAS are associated with disease progression.68,69 CD44v6 is a promising protein involved in cell migration and adhesion. It has been shown that CD44v6 expression is lower in high-grade compared with low-grade lesions, and is associated
with an increased risk of disease recurrence.70 CD44v6 seems to specifically target UCB, as no correlation with other bladder tumor histologies, tumor size, and patient characteristics were observed.71
Hypoxia-Related and Angiogenesis-Related Biomarkers Several studies have shown an association of worse outcomes with altered angiogenesisrelated markers. Shariat and colleagues72 showed that the overexpression of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor, as well as the decreased expression of thrombospondin 1 were associated with higher UCB stages, lymphovascular invasion, and lymph node metastasis. Alterations of VEGF are influenced by the expression of its 2 receptors, VEGF-receptor-1 (VEGFR1) and 2 (VEGFR2). Although VEGFR2 seems to be overexpressed in muscle-invasive bladder cancer, VEGFR1 tends to be overexpressed in NMIBC.73 Finally, a recent meta-analysis of 11 studies supports the hypothesis that overexpression of VEGF in NMIBC is associated with poor prognosis.74 CD34 is a transmembrane cell surface glycoprotein involved in tumor angiogenesis.75 In patients with NMIBC, CD34 microvessel density has been found to be associated with recurrence after Bacillus Calmette-Guerin.76 Another novel prognostic marker involved in angiogenesis is hyaluronoglucosaminidase 1 (HYAL-1). Based on tissue microarrays from a cohort of 178 UCB specimens, Kramer and colleagues77 showed a significant association between HYAL-1 and disease progression to invasive stages. Hypoxia is a stress signal during tumor growth and regulates gene expression. The carbonic anhydrase IX (CAIX) is a hypoxia-related biomarker, responsible of intracellular pH regulation and hydrogen ion (H1) flux.78 Among NMIBC, higher CAIX expression was found to be associated with lower recurrence-free survival and a 6.5-fold higher risk of disease progression.79 The 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 (PFKFB4) is also induced by hypoxia. Yun and colleagues80 found that overexpression of PFKFB4 was significantly associated with highstage UCB and multiple tumors compared with low stage and single tumors (P
KEY POINTS Urothelial carcinoma of the bladder (UCB) is the most expensive solid tumor to treat because of its high recurrence rate and the need of continued cystoscopic surveillance. Cystoscopy has a high sensitivity and specificity, but is a costly and invasive procedure; urinary cytology is noninvasive and highly specific and shows very low overall sensitivity, in particular in low-grade nonmuscle invasive bladder cancer (NMIBC). Several biomarkers may have better sensitivity than voided urinary cytology, but they are not accurate enough to replace cystoscopy and cytology. NMP22, BTA, ImmunoCyt/uCyt1 and UroVysion/FISH have been approved by the Food and Drug Administration for screening and follow-up of patients with NMIBC in combination with cystoscopy. The long interval required for validation, testing, and approval of the assays and the lack of standardization could explain the present failure of biomarkers to predict NMIBC development, recurrence, and progression.
Urothelial carcinoma of the bladder (UCB) is the most expensive solid tumor to treat because of its high recurrence rate and need of continued surveillance, with an estimated annual cost of $3.7 billion in the United States.1 More than
400,000 patients are newly diagnosed with UCB worldwide every year, and more than 150,000 die from the disease.2 The standard evaluation of a suspected UCB consists of cystoscopy and urinary cytology.3 Although cystoscopy has a high sensitivity and
Funding Sources: Development fund of the CHUV-University Hospital, European Urological Scholarship Programme (I. Lucca); Oesterreichische Nationalbank Anniversary Fund (project number: 15363/2013), Medical-Scientific Fund of the Mayor of the City of Vienna (project number: 14069/2014) (T. Klatte); Nil (M. de Martino, S. Shariat). Conflict of Interest: Nil. a Department of Urology, Comprehensive Cancer Center, Vienna General Hospital, Medical University of Vienna, Wa¨hringer Gu¨rtel 18-20, Vienna A-1090, Austria; b Department of Urology, Centre Hospitalier Universitaire Vaudois, Rue du Bugnon 46, Lausanne CH-1010, Switzerland; c Department of Urology, University of Texas Southwestern Medical Center, 1801 Inwood Road, Dallas, TX 75235, USA; d Department of Urology, Weill Cornell Medical College, New York-Presbyterian Hospital, Cornell University, 1300 York Avenue, New York, NY 10065, USA * Corresponding author. Department of Urology, Comprehensive Cancer Center, Vienna General Hospital, Medical University Vienna, Wa¨hringer Gu¨rtel 18-20, Vienna A-1090, Austria. E-mail address: [email protected] Urol Clin N Am 42 (2015) 225–233 http://dx.doi.org/10.1016/j.ucl.2015.01.004 0094-0143/15/$ – see front matter Ó 2015 Elsevier Inc. All rights reserved.
urologic.theclinics.com
INTRODUCTION
226
Lucca et al specificity, it is a costly and invasive procedure, with a risk of complications, such as urinary infections, pain, and bleeding. Urinary cytology has a low overall sensitivity rate of only 33% to 48%, but a higher specificity of 86% to 90% in the detection of nonmuscle invasive bladder cancer (NMIBC).4,5 Its predictive accuracy improves in high-grade tumors to approximately 90%, whereas it does not go beyond 60% in lowgrade NMIBC.6 Urinary cytology after bladder washing seems to have a higher sensitivity than from voided urine.7 In UCB, the TNM staging together with various prognostic models, such as the European Organization for the Research and Treatment of Cancer (EORTC) and Spanish Urological Club for Oncological Treatment (CUETO) scoring systems, allow defining the risk of recurrence and progression after the surgery.3,8 However, among patients with UCB staged similarly, there is still a considerable variation in rates of recurrence, disease progression, and response to treatment. Therefore, biomarkers are needed to improve prognostication. Intense work is being done in the field of NMIBC biomarkers with the main goals of reducing the number of invasive cystoscopic evaluations through early diagnosis and identifying those patients with a higher risk of disease progression. The identification of valuable NMIBC biomarkers would improve our understanding of the biological pathways involved in the genesis and progression of the tumor, but would also add information about UCB outcomes. A biomarker is an indicator of biological and pathogenic processes or pharmaceutical responses to a therapeutic intervention.9 The ideal
Fig. 1. Classification of biomarkers for NMIBC.
biomarker should be noninvasive, rapid to analyze, objective, easy to perform, sensitive, specific, reproducible, and, above all, cost-effective. However, to date, the available biomarkers are all lacking one or more of these characteristics. Several biomarkers showed a higher sensitivity than voided urinary cytology, but they are not accurate enough to replace it. Indeed, although several studies have been published on NMIBC biomarkers in recent years, none of them is recommended for use in clinical practice. The classification of UCB biomarkers is relatively complex. They could be classified according to (1) the sample used for their identification, (2) their objective, (3) the biomolecule, or (4) the type of measurements applied (Fig. 1).10 Because the literature on this topic is vast, the aim of this review was to provide a summary of the most relevant biomarkers investigated in the last period as predictor of NMIBC, recurrence, and aggressiveness.
DIAGNOSTIC BIOMARKERS FOR EARLY DETECTION OF NONMUSCLE INVASIVE BLADDER CANCER Diagnostic biomarkers aim to detect the disease in symptomatic patients or in those having an increased risk, if possible at early stages.11 NMIBC is an attractive tumor for screening because of its well-known risk factors and its favorable outcomes when diagnosed at early stages. Macrohematuria is the most common symptom, but it is not specific and is not related to the pathologic stage. Madeb and Messing12 screened 1575 men aged 50 years or older without a history of UCB,
Biomarkers for Localized Bladder Cancer hematuria, or previous treatment with pelvic radiation therapy for hematuria. Although 6% to 10% of men with hematuria were diagnosed with UCB, hemoglobin dipsticks had a positive predictive value of only 8.1%. It has been hypothesized that a combination of home hematuria urinary testing and a panel of biomarkers may improve the sensitivity and specificity of the screening. The Bladder Cancer Urine Marker Project (BLU-P) tested a total of 1747 Dutch men with a home hematuria test followed by a molecular marker panel, which included nuclear matrix protein 22 (NMP22), microsatellite analysis (MA), the fibroblast growth factor receptor 3 (FGFR3) mutation snapshot assay, and a custom methylationspecific (MLPA) test.13 The specificity and sensitivity of any positive molecular test were 95.9% and 80.0%, respectively. However, of the 71 men who finally underwent cystoscopy, only 4 UCBs and 1 upper tract urothelial carcinoma were detected, showing a lack of benefit of this combined screening approach in a large unselected asymptomatic population.
Protein Biomarkers Several studies have reported the association of urinary protein biomarkers and UCB, leading to the Food and Drug Administration (FDA) approval of NMP22, bladder tumor antigen (BTA), and ImmunoCyt/uCyt1 for screening and follow-up in combination with cystoscopy. NMP22 is a protein found in the nuclear matrix and is involved in replication and gene expression. The sensitivity and specificity of the urinary NMP22 test to detect UCB is approximately 68% and 79%, respectively.14–16 The test is commercially available and detects urinary levels of the nuclear mitotic apparatus protein, which is elevated in the urine of patients with UCB. However, a high rate of false-positive tests has been reported, often due to macrohematuria, urinary tract infections, and instrumentation. Ponsky and colleagues17 showed that benign inflammatory or infectious conditions, renal or bladder calculi, current or recent history of foreign body in the urinary tract, any bowel interposition segment, and other genitourinary cancers are responsible for reducing the positive predictive value and the specificity. Based on an experimental model, Miyake and colleagues18 suggested that urinary NMP22 assays may diagnose UCB by counting the amount of cell turnover as consequences of surface shedding from bladder tumors rather than identifying a specific tumor antigen. That would explain the false-negative NMP22 tests in those with reduced urinary cellularity due to the characteristic of the
tumor or an error in sampling. Moreover, concentrated urine may be also responsible for falsepositive tests because of an overestimation of NMP22 levels.19 Taken together, NMP22 cannot be considered an alternative to cystoscopy. The BTA test detects the human complement factor H-related protein, which seems to prevent cell lysis in UCB.20 The original BTA test (Bard Diagnostics, Redmond, WA) was taken off the market because of its very low specificity and sensitivity.21 Two tests are currently available, the BTA stat and BTA TRAK (Polymedco, Cortlandt Manor, NY). Although the first is qualitative and gives an immediate result, the second is quantitative and requires experienced laboratory processing. The overall BTA sensitivity is higher than urinary cytology (61%–67%), but the specificity seems to be lower (75%–78%).4,22 The presence of blood in the urine interferes with the BTA tests, causing a high rate of false-positive results due to the cross-reactivity to antigens on red blood cells.23 Some investigators have suggested the use of BTA as serum biomarker, measuring complement factor H introduced by bleeding.24 ImmunoCyt/uCyt1 (Scimedx, Denville, NJ) is approved by the FDA for the follow-up of patients with NMIBC. It combines 3 monoclonal antibodies and operates as a second-level test in case of atypical urinary cytology. Odisho and colleagues25 performed ImmunoCyt/uCyt1 tests on 506 voided urine samples with atypical urinary cytology. For low-grade UCB, the sensitivity, specificity, and negative predictive value were 75%, 50%, and 82%, respectively. For high-grade UCB, the values were comparable (74%, 44%, and 79%, respectively). Moreover, Teˆtu and colleagues26 showed a significant improvement of the urinary cytology sensitivity when combined with ImmunoCyt/ uCyt1, with a sensitivity of 78% in Ta, 83% in T1, and 100% in carcinoma in situ.
Chromosome Biomarkers Another FDA-approved urinary biomarker is UroVysion (Abbott Laboratories, Abbott Park, IL), a multitarget fluorescence in situ hybridization (FISH) assay that detects aneuploidy in chromosomes 3, 7, and 17 as well as focal loss of 9p21 containing the p16 tumor-suppressor gene.27 A meta-analysis of 14 studies involving 2477 UroVysion-FISH tests found an overall sensitivity of 86% (82%–89%) for UroVysion compared with 61% (56%–66%) for urinary cytology.28 However, positive FISH did not reach the level of evidence for the initial diagnosis of UCB. In patients with NMIBC with positive cytology but negative cystoscopy, the prediction of disease occurrence or
227
228
Lucca et al recurrence is challenging. In a study investigating 243 patients with NMIBC with atypical urinary cytology, a positive UroVysion-FISH test was a significant predictor of recurrence (hazard ratio 2.35, P 5 .001) in multivariate analysis.29 Finally, Seideman and colleagues30 showed that patients with NMIBC with atypical urinary cytology and positive UroVysion-FISH had a shorter time to recurrence compared with those with a negative UroVysion-FISH (12.6 vs 17.9 months), even when cystoscopy was negative.
DNA Methylation Biomarkers DNA methylation is one of the most studied epigenetic mechanisms, which commonly occurs in CpG-rich regions that are located in the promoter regions and lead to gene inactivation.31 Because CpG hypermethylation does not occur in normal cells, many methylated genes have been linked to UCB pathways.32–38 The detection of methylation biomarkers has several advantages over conventional genetic markers because of their high degree of specificity, which is important when DNA is available in limited quantities (eg, from biological fluids such as urine). Moreover, they play a role in both diagnosis and prognosis of UCB. Su and colleagues39 built a panel of biomarkers, including a combination of a transcription factor (SOX1), a specific LINE-1 element, and interleukin-1 receptorassociated kinase 3 (IRAK3), which could detect a higher rate of early recurrences than urinary cytology and cystoscopy. DNA methylated biomarkers, such as OTX1, ONECUT2, and OSR1, have been studied in combination with FGFR3 assays, showing a sensitivity of 79% and a specificity of 77% in patients with NMIBC.40 Because DNA methylation is involved in UCB pathogenesis, numerous promoters of DNA hypermethylation are being investigated (ie, tumorsuppressor genes, proto-oncogenes, genes involved in cell adhesion, and genes regulating the cell cycle).41,42 In a recent study, histone deacetylases (HDAC)-1 and HDAC-2 overexpression was associated with the development of high-grade NMIBC, suggesting a potential role of HDAC inhibitors as chemotherapeutics.43 Moreover, downregulated KLF4 has been found to promote DNA hypermethylation and may be associated with a higher risk of early recurrence and disease progression.44
Micro RNA Biomarkers Micro RNAs (miRNAs) are short noncoding RNA molecules of 18 to 22 nucleotides that are able to modulate posttranscriptional protein expression. Upregulation or downregulation of
miRNAs may influence UCB tumorigenesis by acting as oncogenes or tumor-suppressor genes. A total of 2216 miRNAs have been registered so far.45 Most miRNA studies focus on tissue samples, whereas only one-third is on urine samples. Low-grade UCB is characterized by miRNA downregulation, whereas in high-grade UCB, there is typically miRNA upregulation.46 A recent metaanalysis identified 15 miRNAs as the most promising diagnostic and prognostic UCB biomarkers, of which 11 were upregulated (miRNA-96, miRNA-138, miRNA-126, miRNA-182, miRNA143, miRNA-222, miRNA-21, miRNA-133b, miRNA-518c-5p, miRNA-452, and miRNA-129) and 4 were downregulated (miRNA-200c, miRNA-99a, miRNA-100, and miRNA-29c).47 Among them, miRNA-99a expression was inversely correlated with the aggressiveness of UCB, suggesting a role of miRNA-99a as tumor suppressor.48 Also miRNA-100 might act as a tumor suppressor. Indeed, when artificially reintroduced in tumor tissue, it is able to suppress cell proliferation and motility, inducing cell-cycle arrest in vitro.49 Finally, some promising miRNAs, such as miRNA-141 and miRNA-639, have been recently studied in plasma, but final results showed no differences in serum levels between patients with UCB and controls.50
Messenger RNA Biomarkers Survivin, also known as BIRC5 or EPR-1, has a pivotal role in promoting cell proliferation and preventing apoptosis. Its sensitivity and specificity to predict UCB have been estimated at approximately 76% to 78% and 82% to 95%, respectively, and when survivin tests are combined with urinary cytology, the accuracy increases significantly.51–53 Moreover, patients with NMIBC expressing high survivin levels have shown a lower risk of disease recurrence.54 Another interesting mRNA biomarker is Aurora A kinase (AURKA), a cell-cycle–associated serine/threonine kinase involved in genetic stability. The AURKA overexpression seems to be specific of UCB tissue and in voided urine samples may be associated with higher UCB stage and grade, and worse survival.6
PROGNOSTIC BIOMARKERS IN NONMUSCLE INVASIVE BLADDER CANCER In NMIBC, presence of carcinoma in situ, tumor stage and grade, tumor size, and numbers of lesions are established variables to predict disease recurrence and progression. It has been estimated that 10% to 20% of patients progress to a muscle-invasive stage.55,56 New prognostic biomarkers that identify patients with a high risk
Biomarkers for Localized Bladder Cancer of progression, and may therefore be candidates for a more aggressive therapy, are needed.
Inflammatory Biomarkers Inflammation plays a pivotal role in the development and progression of UCB.57,58 It is commonly accepted that the tumor microenvironment triggers the release of cytokines, subsequently influencing the systemic inflammatory response.59 Measurements of systemic inflammatory response represent valuable, reproducible, and inexpensive prognostic biomarkers. According to a recent meta-analysis, inflammatory cells, UCB-costimulatory molecules, serum cytokines, and the C-reactive protein were all significantly associated with survival.60 Some inflammatory measurements have been combined in ratios to improve their prognostic value. A high neutrophil-to-lymphocyte ratio (NLR) has been associated with poor survival and higher risk of disease recurrence in NMIBC. Mano and colleagues61 estimated a 3-year progression-free survival rate of 61% in patients with NMIBC with a preoperative NLR greater than 2.41, compared with a rate of 84% in patients with an NLR of 2.41 or less (P 5 .004).
Cell Cycle and Adhesion Markers In the past decade, several molecular biomarkers have been identified as potential predictors of UCB progression to an invasive stage.62 The p53 protein is a well-known tumor-suppressor protein with a crucial role in the prevention of genomic mutations. Besides its role in predicting early recurrence, several studies showed an association between p53 overexpression and UCB progression.63,64 Shariat and colleagues65 studied p53 in combination with other molecular markers (pRB, p21, and p27). The investigators found that a high number of altered biomarkers was significantly associated with an increased risk of disease recurrence (P 5 .011) and progression (P 5 .005). Deletions of the long arm of chromosome 9 and mutations in FGFR3, H-RAS, and PI3K also may influence the development of UCB.66 Lindgren and colleagues67 found higher frequency of FGFR3 mutations in low-grade NMIBC (80%) than in high-grade NMIBC (less than 10%), suggesting a protection role of FGFR3 against disease progression. On the other hand, others failed to show that deletions of the long arm of chromosome 9 and mutations of HRAS are associated with disease progression.68,69 CD44v6 is a promising protein involved in cell migration and adhesion. It has been shown that CD44v6 expression is lower in high-grade compared with low-grade lesions, and is associated
with an increased risk of disease recurrence.70 CD44v6 seems to specifically target UCB, as no correlation with other bladder tumor histologies, tumor size, and patient characteristics were observed.71
Hypoxia-Related and Angiogenesis-Related Biomarkers Several studies have shown an association of worse outcomes with altered angiogenesisrelated markers. Shariat and colleagues72 showed that the overexpression of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor, as well as the decreased expression of thrombospondin 1 were associated with higher UCB stages, lymphovascular invasion, and lymph node metastasis. Alterations of VEGF are influenced by the expression of its 2 receptors, VEGF-receptor-1 (VEGFR1) and 2 (VEGFR2). Although VEGFR2 seems to be overexpressed in muscle-invasive bladder cancer, VEGFR1 tends to be overexpressed in NMIBC.73 Finally, a recent meta-analysis of 11 studies supports the hypothesis that overexpression of VEGF in NMIBC is associated with poor prognosis.74 CD34 is a transmembrane cell surface glycoprotein involved in tumor angiogenesis.75 In patients with NMIBC, CD34 microvessel density has been found to be associated with recurrence after Bacillus Calmette-Guerin.76 Another novel prognostic marker involved in angiogenesis is hyaluronoglucosaminidase 1 (HYAL-1). Based on tissue microarrays from a cohort of 178 UCB specimens, Kramer and colleagues77 showed a significant association between HYAL-1 and disease progression to invasive stages. Hypoxia is a stress signal during tumor growth and regulates gene expression. The carbonic anhydrase IX (CAIX) is a hypoxia-related biomarker, responsible of intracellular pH regulation and hydrogen ion (H1) flux.78 Among NMIBC, higher CAIX expression was found to be associated with lower recurrence-free survival and a 6.5-fold higher risk of disease progression.79 The 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 (PFKFB4) is also induced by hypoxia. Yun and colleagues80 found that overexpression of PFKFB4 was significantly associated with highstage UCB and multiple tumors compared with low stage and single tumors (P
