Potential of urinary biomarkers in early bladder cancer diagnosis

Perspective For reprint orders, please contact [email protected]

Potential of urinary biomarkers in early bladder cancer diagnosis Thomas Lam and Ghulam Nabi †

References

Carcinoma of urinary bladder ranks among the top ten most common cancers worldwide. Approximately 80% of the disease is superficial (limited to mucosa and lamina propria) at the time of presentation. However, the majority of these tumors recur and 15–20% progress into muscle-invasive disease. Cystoscopic surveillance of the urinary bladder remains the standard of care to identify these recurrences on follow-up. Not only is this an invasive procedure, but the sensitivity of cystoscopy can be as low as 70%, so there can be up to 30% of tumors that are missed. Urinary cytology, with recognized limitations, has been used as an adjunct to this procedure, pending discovery of alternate urinary biomarkers. In the past decade there has been tremendous advancement in producing urinary biomarkers for urinary bladder cancer research, reflecting advancements in genomics and proteomics. An ideal biomarker should be able to replace cystoscopic examination and be cost effective. Unfortunately, most of the identified protein or molecular biomarkers have failed this test. This article critically appraises the status of these urinary biomarkers in urinary bladder cancer, in addition to highlighting some of the difficulties in this research area.

Affiliations

Expert Rev. Anticancer Ther. 7(8), 1105–1115 (2007)

CONTENTS Identification of suitable urinary biomarkers Expert commentary Five-year view Key issues Financial disclosure

†

Author for correspondence University of Aberdeen, Academic Urology Unit, First floor, Health Sciences Building, Aberdeen AB25 2ZD, UK Tel.: +44 122 455 4877 Fax: +44 122 455 4580 [email protected] KEYWORDS: early bladder cancer diagnosis, urinary biomarkers

www.future-drugs.com

Bladder cancer is a common urological cancer, with more than 60,000 new cases diagnosed and accounting for large numbers of deaths due to cancers in 2006 [1]. Depending upon the level of tumor invasion into the bladder wall, bladder cancer can be categorized into superficial (invasion up to lamina propria) or muscle-invasive disease (invasion into detrusor muscles). Up to 80% of all bladder cancer at presentation is superficial disease, which can be managed by transurethral resection. However, because of a higher tendency to recur (∼50–90%) despite a complete initial endoscopical removal, follow-up cystoscopic surveillance is required [2]. This is not only an invasive method, but also makes bladder cancer care a less cost-effective method for any healthcare system. Bladder cancer is currently the most expensive of all cancers to manage in terms of costs per patient [3]. Cystoscopy remains the gold standard for detecting new bladder cancers. However, up to 10% of all cancers can be missed, especially if they are flat lesions or carcinoma in situ [4–6].

10.1586/14737140.7.8.1105

Moreover, the procedure is invasive, uncomfortable for patients and carries a risk of complications. Urinary cytology is used as an adjunct to cystoscopy. However, urine cytological analysis for malignant cells has a number of recognized limitations [7–9]: • Overall low sensitivity • High subjectivity • Significant interobserver variability • Requirement of specialized laboratory personnel Consequently, there is an urgent need to develop novel biomarkers that could significantly improve the diagnosis of bladder cancer, especially in its early stages. It is crucial to understand the clinical care pathways of superficial bladder cancer in contemporary urological practice. FIGURE 1 illustrates various pathways of diagnosis and follow-up of bladder cancer. Specifically, we need urgent biomarkers to diagnose, prognosticate and reduce the frequency of surveillance cystoscopic examination or, preferably, replace it

© 2007 Future Drugs Ltd

ISSN 1473-7140

1105

1106

Less common route

Most common route

Flexible cystoscopy

Bladder tumor seen on flexible cystoscopy or bladder tumor suspected (red patch, abnormal area)

Wait for histology report

Post-TURBT CT scans

Very large tumor and suspicion of muscle invasion on cystoscopy and examination under anesthesia

Tumor is large (>5 cm)

Pre-TURBT CT-scan of abdomen and pelvis

Single intravesical dose of mitomycin (40 mg) and removal of catheter at postoperative day 1–2 Discharge from hospital

Operative note of: – Size of tumor, site of tumor in bladder and findings of examination under anesthesia – Appearance of rest of the bladder

Planned admission of TURBT or bladder biopsies for suspicious-looking areas under general or spinal anesthesia

Flexible cystoscopy, urinary cytology for malignant cells, ultrasound of abdomen and intravenous urogram

Diagnosis and prognosis of bladder tumors will improve with biomarkers

Frank or microscopic hematuria

Figure 1A. Diagnosis and follow-up of bladder cancer (care pathways). GP: General practitioner; TURB: Transurethral resection of bladder tumor.

Persistent irritative lower urinary tract symptoms or recurrent urinary tract infections in female patients seen in clinic

• GP referral (most common route) • Emergency admission with clot retention

Lam & Nabi

Expert Rev. Anticancer Ther. 7(8), (2007)


Recurrences

Discuss MDT for radical treatment

Adjuvant chemotherapy (mitomycin/BCG) and cystoscopy 3 monthly for first 2 years and yearly thereafter

3-monthly cystoscopy for first year, 6-monthly for second year and yearly afterwards, if recurrence 1. Biopsy and cystodiathermy 2. Cystodiathermy + Single dose of mitomycin

First cystoscopy at 9 months and then annual cystoscopy (flexible under local) Life-long or 5 years (provided no further recurrences) depending on local follow-up policy

Continue endoscopic control: 1. Biopsy and cystodiathermy 2. Cystodiathermy + Single dose of mitomycin

90% chance of recurrence at 1 year

Biomarkers are expected to reduce the number of follow-up cystoscopies

High risk (10%) – Multiple tumors at presentation and recurrence at 3 months



Figure 1B. Diagnosis and follow-up of bladder cancer (care pathways). BCG: Bacillus Calmette-Guerin; MDT: Multidrug therapy.

Cystoscopy at 3 months (pTa–T1 tumors)

Intermediate risk (30%) – Solitary tumor at presentation and recurrence at 3 months – Multiple tumors at presentation and no recurrence at 3 months

Low-risk group (60%) – Solitary tumor at presentation and no recurrence at 3 months

Urinary biomarkers in early bladder cancer diagnosis

1107

Lam & Nabi

with a noninvasive test. FIGURES 1A & B illustrate these areas highlighted in the care pathways. We presume that the cost of bladder cancer management, especially surveillance, is an important factor driving a continuous search for ideal biomarker/s. Urine is an ideal and rich source of biomarkers for various reasons: • Easy and noninvasive means of obtaining samples; • Consistent and continuous contact of urine with various urological organs; • Shedding of cells, proteins, enzymes, nucleic acids and metabolic products into urine during various physiological and pathological processes. While the clonal origin of bladder cancer remains controversial, there is compelling evidence to support the oligoclonal and metachronous nature of these tumors, and this has implications for their diagnosis and treatment [10]. In particular, it paves the way for the development of molecularly targeted agents that can be utilized for diagnosis and therapy in bladder cancer. In this regard, multiple changes at a molecular level have been described and several of these have been exploited to enable the development of urine-based tests to diagnose bladder cancer. Recent years have seen a discovery of various novel urinary biomarkers, which are summarized in TABLE 1. Furthermore, their current status and perceived advantages and disadvantages are also described. Identification of suitable urinary biomarkers

The developmental process of novel biomarkers from the laboratory to the clinical practice is illustrated in FIGURE 2. It follows an orderly process. The first phase of biomarker development is the identification of suitable candidates. This often begins with preclinical studies, comparing tissue samples, urine or serum of patients with bladder cancer with those from healthy subjects. The purpose of this exploratory phase is to identify molecular characteristics that are unique to bladder cancer in order to develop a test that detects the disease. Historically, the identification of biomarkers has depended on traditional laboratory techniques, such as immunohistochemistry and western blotting. Although reliable, these techniques are slow, tedious and labor intensive. However, advances in molecular biology have fuelled new techniques, which enable the analysis of huge repertoires of novel candidate markers quickly, efficiently and in an automated fashion. These include three main techniques, genomics, proteomics and metabonomics, which essentially serve to profile the molecular or genetic signatures of specific diseases in a high-throughput fashion. Genomics is defined as the measurement of gene expression from available sequence information and the expression profile reflects the function and phenotype of a cell. Landmark advances in molecular biology, such as cDNA and oligonucleotide microarrays on chips [11], serial analysis of gene expression (SAGE) [12] and the development of assaying techniques in miniaturized platforms have accelerated the development of genomics. Complementary DNA and oligonucleotide

1108

microarrays enable the simultaneous assessment of the transcription of tens of thousands of genes and their relative expression between normal and cancerous cells. Extensive SAGE libraries have enabled the formation of databases that may have immediate application in studies of diagnosis and prognostic assessment. Analysis of global gene and protein profiles will help to detect patterns of gene expression and protein modification, and aid understanding of the unique pathology of specific cancers. However, the field of microarrays is still in its infancy, hence problems such as inconsistency, lack of robustness and lack of concordance among microarray data have been encountered [13]. Proteomics is a research field focused on characterizing the molecular and cellular dynamics in protein expression and function on a global scale. Proteomic methods identify the functioning units of expressed genes through the analysis of cellular proteins in order to provide a protein fingerprint [14,15]. The proteome reflects the intrinsic genetic program of the cell and the impact of its immediate environment, and hence it can be a valuable resource in biomarker discovery. Progress in the field of mass spectrometry has complemented traditional protein chemistry methods in the identification of novel proteins. Through the use of matrix-assisted laser desorption/ionization (MALDI) and tandem mass spectroscopy, mass-spectral analysis is helping in the detection and identification of proteins and sequence analysis from biological fluids, such as urine. Protein chips using surface-enhanced laser desorption/ionization (SELDI) are being developed as high-throughput assays for a panel of markers in determining protein fingerprints [16]. In addition, chip technology that can characterize proteins from complex biological fluids at extremely low concentrations is particularly promising, and is likely to contribute significantly to the discovery of novel biomarkers. One of the main advantages of proteomics is that it enables the analysis of an entire proteome or subproteome in a single experiment, and this allows the cross-referencing of protein alterations with a corresponding disease state at a given time point. Another advantage over genomics is that proteomics considers proteins, which are ultimately responsible for the disease phenotype, whereas genomics is mainly concerned with genetic expression. In addition, proteomics can identify alterations in post-translational modifications, cellular trafficking and even total expression levels, all of which may not be detected by RNA-based genomics studies. It is important to note that post-translational modifications of proteins, which are not detected through RNA analysis, may occur at different stages of tumor development; this phenomenon can be investigated via proteomics-based studies. Finally, since most US FDA-approved diagnostic tests are protein based, findings from proteomic studies can be used to develop clinically useful tests. Traditionally, proteomic studies have been used for biomarker discovery, and the clinical test is typically an enzyme-linked immunosorbent assay (ELISA)based assay. However, with the advent of high-throughput protein assays, some of these analytic instruments may, at some stage, be applied as proteomics-based clinical diagnostic tests.



Table 1. Current status of diagnostic urinary biomarkers for bladder cancer. Name of biomarker

Test name

Advantages

Disadvantages

Soluble urinary proteins Human complement BTAstat; BTA-TRAK (both Sensitivity 50–70% factor H-related protein approved by the US FDA (bladder tumor antigen) for surveillance only)

Specificity 50–75%; false positives with hematuria, proteinuria, infection, stones, inflammation and some foods or drugs

Soluble Fas

In development

Preliminary study suggests higher specificity than NMP-22

Previously assessed in surveillance only; not yet commercially available

NMP-22

NMP-22 BladderCheck test® (approved by FDA for surveillance and screening)

Inexpensive, rapid and operator independent

Wide-ranging sensitivity (69–89%) and specificity (65–91%); false positives with infection, inflammation, stones, hematuria and post cystoscopy

BLCA4 (nuclear transcription factor)

In development

Preliminary studies suggest 95% sensitivity and 89% specificity

Currently being evaluated in multicenter trial

Survivin (antiapoptotic protein)

In development

Preliminary studies suggest higher High levels of prostatic secretion of survivin sensitivity and specificity than NMP-22 represent confounding factor and cytology

uPA (serine protease)

In development

Preliminary study suggests higher specificity than NMP-22 and cytology

May need to be integrated with other biomarkers and cytology to be clinically useful

FDP (clotting cascade protein)

Accu-Dx (FDP test; approved by FDA for surveillance only)

86% specificity; rapid

Sensitivity 68%; false positives with hematuria; not commercially available because product withdrawn by manufacturer

Hyaluronic acid and hyaluronidase (extracellular matrix proteins)

In development

94% specificity for recurrence; up to 85% sensitivity; detects low-grade and low-stage disease

Currently being evaluated in multicenter trial

Mcm5p

In development

High sensitiviy and specificity; sensitivity and specificity can be varied by applying different assay cut points

Assessed in one study only

Centromeres on chromosomes 3, 7, 17 and 9p21

UroVysionTM FISH (approved by FDA for primary diagnosis and surveillance)

84% sensitivity; 95% specificity

Expensive; requires trained personnel; thresholds for positive test remain arbitrary; requires intact urothelial cells; adjunct to conventional cytology

Bladder cancer variant of CEA and bladder cancer mucins

ImmunoCytTM (approved 86% sensitivity by FDA for surveillance only in conjunction with cytology)

79% specificity; fluorescence interpretation can be difficult; requires intact urothelial cells; adjunct to conventional cytology

Lewis X antigen (blood group antigen)

In development


Expression of Lewis X antigen by benign umbrella cells may interfere with test; requires intact urothelial cells; adjunct to conventional cytology

Cytokeratins 8, 18, 19, 20 (cytoskeletal proteins)

UBC-Rapid; UBC-ELISA; CYFRA21–1 (ES700)

>90% sensitivity; >80% specificity

False positives with factors such as infection or stones; requires intact urothelial cells; adjunct to conventional cytology

Cell nucleus and DNA

Quanticyt

Combines DNA ploidy with nuclear morphometry; automated system

Need for bladder wash-out sample rather than voided urine; requires intact urothelial cells; adjunct to conventional cytology

Cell-based biomarkers

BTA: Bladder tumor antigen; CEA: Carcinoembryonic antigen; FDP: Fibrin degradation products; FISH: Fluorescent in situ hybridization; NMP: Nuclear matrix protein; PSCA: Prostate stem cell antigen; TSG: Tumor-suppressor gene; uPA: Urokinase-type plasminogen activator. Modified and adapted from [19,21,23,30].


1109

Lam & Nabi

Table 1. Current status of diagnostic urinary biomarkers for bladder cancer (cont.). Name of biomarker

Test name

Advantages

Disadvantages

Mucin 7 (mucosal glycoprotein)

In development

87% specificity

Low sensitivity (68%); only assessed in one study to date; requires intact urothelial cells

PSCA (surface antigen)

In development


Assessed in one study only; adjunct to conventional cytology

DNA microsatellites (detection of genomic instability)

In development

High sensitivity and specificity (approaching 90%) independent of grade and stage

Labor intensive; expensive; lack of methodological standardization; currently being evaluated in multicenter trial

Telomerase (RNA enzyme conferring cellular immortality)

In development

High sensitivity (90%) and specificity (88%) in screened population

Low sensitivity and specificity for surveillance; poor reproducibility; stability of telomerase is variable; false positives with factors such as infection or stones

TSG (detection of DNA hypermethylation)

In development

96% specificity; 82% sensitivity

Assessed in one study only

FGFR3

In development

89% specificity

62% sensitivity; assessed in one study only, involving recurrent tumors

Centrosomal abnormalities

In development

100% specificity; 89% sensitivity

Assessed in one study only; requires bladder wash-out specimen

Nucleic acid biomarkers

BTA: Bladder tumor antigen; CEA: Carcinoembryonic antigen; FDP: Fibrin degradation products; FISH: Fluorescent in situ hybridization; NMP: Nuclear matrix protein; PSCA: Prostate stem cell antigen; TSG: Tumor-suppressor gene; uPA: Urokinase-type plasminogen activator. Modified and adapted from [19,21,23,30].

Metabonomics refers to the systematic and thorough profiling of metabolite levels and their temporal changes as influenced by intrinsic and extrinsic factors. This is achieved through the use of novel technologies, such as nuclear magnetic resonance spectroscopy and mass spectrometry. Such metabolic profiling requires the evaluation of complex multidimensional spectroscopic data sets using multivariate statistical tools. The strengths and weaknesses of this approach have been summarized elsewhere [17]. The same authors also argue, with some merit it has to be said, that in some instances it might be appropriate to integrate information at the genomic, proteomic and metabonomic levels in order to maximize the probability of observing meaningful and relevant biochemical changes in pharmaceutical research and development, in spite of these apparently different and contrasting approaches. Clinical development of biomarker assays

Once the appropriate urinary biomarker has been identified, the next step is to develop the biomarker into a clinically useful assay. The systematic and step-wise development of a biomarker has been the subject of much debate and controversy, and has been discussed in detail previously [18–20]. In brief, the suggested phases of development are as follows: Phase I: preclinical exploratory studies to identify suitable biomarkers followed by development of an assay based on tumor tissue samples

The assay is assessed on tumor tissue obtained from patients with bladder cancer and on normal bladder tissue from

1110

matched controls. The primary aims are to identify and prioritize leads for potentially useful biomarkers. Some types of biomarkers, for instance those identified directly from urine rather than from tumor tissue, can leave this phase and proceed straight through to Phase II. Phase II: development of a clinically useful assay based upon a specimen that can be obtained noninvasively (e.g., unrine)

The assay is assessed retrospectively on urine specimens obtained from patients with bladder cancer and matched controls. However, the study and control groups must be carefully selected to reflect the target population in clinical practice. The primary aims are to determine the true-positive rate, false-positive rate and receiver operating characteristic curve for the assay, and to assess its ability to distinguish patients with bladder cancer from those without it. Phase III: londitudinal repository studies to provide evidence regarding the efficacy of the biomarker to detect early-stage bladder cancer

This is performed on repositories of clinical specimens collected and stored from a cohort of patients with early bladder cancer and from matched controls. As for Phase II, the study and control populations should reflect the target population in clinical practice. The primary aims are to evaluate the capacity of the biomarker to diagnose early bladder cancer and to define the criteria for a positive test in preparation for Phase IV.



Phases

1

High throughput

Low throughput

Immunohistochemistry western blotting

Biomarker Protein identified

Genomics Proteomics Metabolomics

technologies for early cancer detection [101]. The above five-phase model of biomarker development was pioneered by the EDRN. Critical review & classification of diagnostic biomarkers in bladder cancer

The biochemical measurement of soluble biomarkers in the urine provides a diagnostic means that can either complement existing methods (i.e., cystoscopy and cytology) or be an independent mode of Test diagnostic accuracy of new 3 diagnosis and surveillance. Due to biomarker compared with ‘gold standard’ advances in the field of molecular biology, (sensitivity and specificity) biomarkers involved in different stages of the bladder cancer cell evolutionary path4 way have been identified and targeted for Population-based studies – Case control development as diagnostic tools. They can – Randomized controlled trials be classified into different categories, – Cost–effectiveness including tumor-associated surface antigens, blood-group antigens, growth factors, proteins involved in the cell cycle or 5 Assess impact of biomarker on apoptotic pathways, extracellular matrix disease outcomes (mortality and proteins, clotting cascade, cytosolic prosurvival in cancers) teins, nucleic acids and chromosomes. From a methodological perspective, uriFigure 2. Identification of biomarkers (journey through different phases). nary markers fall into a few broad groups: Phase IV: Prospective studies to determine the operating soluble urinary proteins, cell-based biomarkers and nucleic acid characteristics of the biomarker assay in a relevant population biomarkers. As a complete review of each specific biomarker is by establishing the true-positive, true-negative, false-positive beyond the scope of this article, the following is a brief sum& false-negative rates mary of the relative merits and problems associated with each The test is applied prospectively to a study population at risk of group of biomarker and a short commentary on some of the developing bladder cancer (e.g., patients presenting with hema- most promising candidate markers currently in development. turia or those with previous history of bladder cancer undergoing surveillance) and patients will also undergo definitive Soluble urinary proteins testing (e.g., with cystoscopy and/or urine cytology) to contruct As illustrated in TABLE 1, soluble urinary proteins are probably a contingency table to determine the above parameters. the most common type of biomarker at all stages of development. The main advantage of assays that detect such biomarkPhase V: cancer control studies to determine whether the test ers is that they are relatively inexpensive, can be performed reduces the cancer burden on the population quickly and do not require the presence of intact urothelial These studies are essentially randomized controlled trials involv- cells in the urine. Three biomarkers have been approved by the ing patients at risk of developing bladder cancer, and patients US FDA for use in either surveillance (i.e., nuclear matrix proare randomized to undergo either the biomarker test or standard tein [NMP]-22, bladder tumor antigen [BTA] and FDP) or investigations (e.g., cystoscopy). The primary aims are to esti- screening (i.e., NMP-22) of bladder cancer. Of these, the mate the reductions in cancer-specific and overall mortality, as NMP-22 test is the most established and rigorously assessed, and it arguably represents one of the best FDA-approved well as to determine the impact on costs and quality of life. biomarkers currently available [21]. However, it also has some Summary inherent limitations that limit its efficacy, as shown in TABLE 1. A systematic, well-planned and concerted effort is essential in Among the non-FDA-approved soluble biomarkers, the most order to develop a biomarker successfully from the laboratory promising is arguably hyaluronic acid and hyaluronidase. to the clinic. In line with this, the US National Cancer Insti- Hyaluronic acid is a free, nonsulfated glycosaminoglycan nortute’s Early Detection Research Network (EDRN) was set up in mally present in body fluids, connective tissues and extracellu2000 to develop and evaluate novel biomarkers for the early lar matrix. It is degraded by hyaluronidase. Levels of detection of cancer and to foster collaboration between differ- hyaluronic acid are elevated in bladder cancer, and hyaluronic ent professionals and groups with an interest in developing new acid fragments and hyaluronidase have been found in the urine 2


Development of antibody-based assay (e.g., ELISA)

1111

Lam & Nabi

of bladder cancer patients [22]. A unique feature of this assay is that it has a particularly high sensitivity for low-grade and lowstage disease [23]. This biomarker is currently being evaluated in multicenter trials. Cell-based biomarkers

Cell-based assays have the main disadvantage of requiring intact urothelial cells in the urinary sediment for their performance. There is reliance upon malignant cells being shed into the voided urinary specimen, and the quality and quantity of such cells can be variable. As such, the specimen can sometimes be of poor diagnostic quality. In addition, the analysis of cells can be labor intensive, time consuming and expensive. Most cell-based biomarkers are intended as an adjunct to urine cytology. In this group, two tests have been FDAapproved for surveillance (i.e., UroVysionTM FISH and ImmunoCystTM) and primary diagnosis (i.e., UroVysion FISH). A recent press release from Diagnocure Inc. suggests that ImmonoCyst has been withdrawn from the market. For the remaining cell-based biomarkers, the most promising are arguably Lewis antigen X and CYFRA21-1 (TABLE 1). Lewis X antigen is a blood-group antigen present only on umbrella cells within the bladder epithelium and it is expressed in bladder cancer, independent of the secretory status of the individual and the cancer grade or stage [23]. CYFRA21-1 is an assay that detects fragments of cytokeratin 19 using two monoclonal antibodies [21]. Further large comparative multicenter studies are required to assess the impact of these new biomarkers in clinical practice.

Assays measuring urinary telomerase activity in bladder cancer have been shown to have high sensitivity and specificity as a primary diagnostic tool [26], although studies on telomerase as a surveillance tool have revealed poor sensitivity and specificity [21]. These contrasting results may reflect some of the methodological difficulties in assessing the performance of biomarkers, especially when they have been applied to different population groups [27]. Moreover, the variable preservation of RNA causes poor reproducibility of the telomerase assays (TRAP). Also, some of the false positives can be derived from leukocytes that can be present in urinary tract infections, which is very common in the same patients that are being screened for bladder cancer. Challenges in the development of urinary biomarkers

In developing a urinary marker, several key issues must be addressed. This section summarizes and discusses issues that we feel are important in the different phases of development of a urinary biomarker. Important attributes of an ideal biomarker assay

The criteria that an ideal biomarker should fulfil are: • Technically simple and quick • Noninvasive • Inexpensive • Reproducible and reliable • High level of accuracy • Well-defined target and control population

Nucleic acid biomarkers

• Acceptable to patients and physicians

The last two decades have seen an exponential growth in the field of genetic engineering. Such progress has paved the way for the identification of novel genes and the development of techniques that exploit genes and nucleic acid sequences for diagnostic and therapeutic purposes. Several nucleic acid biomarkers have been developed to diagnose bladder cancer (TABLE 1). Except for FISH (UroVysion), a DNA-based test, none have received FDA approval. The most promising biomarkers from this group are arguably DNA microsatellites and telomerase. DNA microsatellites are polymorphic repeats scattered throughout the human genome and microsatellite markers (e.g., microsatellite instability or loss of heterozygosity) can isolate genetic mutations associated with bladder cancer [24]. Several studies have demonstrated microsatellite analysis is an efficacious biomarker with high sensitivity and specificity, independent of tumor grade and stage, and it is currently undergoing a large-scale evaluation [21]. Telomerase is an RNA enzyme found in germ cells and cancers, and it helps maintain the length of telomeres, which are DNA sequences on the ends of chromosomes. Telomeres confer genetic stability and are shortened with each successive cell division. Eventually, telomeres are lost, resulting in cell senescence. Bladder cancer cells express telomerase, which prevents loss of telomeres, hence resulting in cellular immortality [25].

A technically simple and quick test is useful because it enables a diagnosis to be made immediately, for example, in a physician’s office. Several point-of-care (POC) assays for the detection of bladder cancer (e.g., NMP-22 POC assay and BTA-Stat POC test) are already in clinical use, and their simplicity and rapidity of obtaining a result are among their strongest features. The benefits of a test that is noninvasive are evident, as the ideal biomarker can potentially replace cystoscopy, which is an invasive and uncomfortable procedure carrying a risk of complications. An efficacious test that is noninvasive is likely to be better accepted by patients as an alternative to cystoscopy for surveillance [28]. Any new test must also be economically justifiable. A reproducible and reliable test is one that displays negligible or very little variability. This variability may have different sources, such as the stability of the biomarker in voided urine, conditions of specimen storage and stability of test reagents. There are two types of variability, namely intra-assay (i.e., within the same patient) or interassay (i.e., between patients), and they can be determined by statistical methods. The test must also have high accuracy, which is a function of sensitivity and specificity. The ideal test should have 100% sensitivity and specificity. However, it is important to note that sensitivity and specificity depend significantly on the target

1112



population. For instance, the specificity of a test may be high when applied to the general population. Nevertheless, when the test is applied to patients at risk of bladder cancer, where there is also a higher prevalence of co-existing benign urological conditions (e.g., infection, inflammation and hematuria), the specificity can be significantly lower, due to the presence of confounding variables within such groups of patients. These issues highlight the importance of judicious selection of both the target and control populations in assessing novel biomarkers, so that the use of the markers in clinical practice is reflected. Another related issue, which affects accuracy, is determining the cut-off value for a positive or negative test. This is important because cut-off values have inverse and proportional relationships to sensitivity and specificity respectively. For instance, reducing the cut-off value can increase the sensitivity of a test (i.e., fewer false negatives), but it also reduces the specificity (i.e., more false positives), while the reverse applies if the cut-off is raised. Consequently, the sensitivity and specificity of a test can be varied by altering the cut-off value, thus the test can be applied to the appropriate target population. For instance, if the marker is to be applied as a screening tool (i.e., primary diagnosis of bladder cancer in a population at risk), a high specificity is important, since the prevalence of the condition is low. In contrast, if the marker is intended to detect tumor recurrences, then a high sensitivity is critical since a missed tumor recurrence can have serious consequences. There is a need to balance such competing demands in order for a test to fulfill its purpose, and this can be performed via statistical methods (e.g., calculation of receiveroperating characteristic curves), as was demonstrated in an elegant article by Stoeber and colleagues [29]. A marker’s performance is also influenced by its positive and negative predictive values, which in turn rely on its sensitivity and specificity. However, it is important to note that both parameters are also dependent on the prevalence of the disease in a given population. Hence, even if a marker displays excellent sensitivity and specificity in different study populations, its positive and negative predictive values may vary among different populations depending on the disease prevalence. These factors make it essential to define and identify the study population appropriately when assessing the performance of the marker in clinical trials. Finally, the test must be acceptable to both patients and physicians in order to be clinically applicable. Patient education (e.g., through public health campaigns) regarding the nature, purpose and efficacy of such diagnostic tests may be important in increasing their acceptability. In practice, no biomarker is likely to possess all of these attributes simultaneously, but these criteria should serve as an essential guide in the development of any biomarker.

Advancements in proteomics research have provided a new window of opportunity to enhance our understanding of biomarkers research and it has already started making impact. In our view, none of the markers available today are sufficiently sensitive or specific to replace the combination of cystoscopy and cytology for the diagnosis of early bladder cancer. In current practice the majority of urinary biomarkers for bladder cancer are used in combination with cystoscopy and/or cytology. Further large multicenter studies with rigorous protocols are required to determine whether a combination of biomarkers with standard approach is able to reduce the frequency of invasive procedures, such as cystoscopy. An increased realization that no single biomarker can replace standard care interventions has made us to consider a combinatorial approach, whereby a panel of different markers, with each targeting a different molecular pathway of carcinogenesis, is used to exploit the heterogeneous nature of cancer. The assay may consist of markers that detect a myriad of molecules, including nucleic acids, proteins, amino acids or metabolic products, and this will not only detect bladder cancer, but perhaps also provide crucial information related to the disease, such as grade and stage of disease, prognosis, phenotypic variant, or even individualized cancer targets that are amenable to molecular therapy. Finally, some of the basic issues, such as impact of biomarker discovery on patient-related outcomes, morbidity, satisfaction, quality of life and survival, must be studied. Furthermore, the cost–effectiveness of any noninvasive test, such as a biomarker must be assessed in comparison with the standard management in any healthcare system. Five-year view

As our knowledge and advancements in urinary proteomics research evolve in future, we should identify the key molecular drivers of recurrences and progression of bladder cancer. We envisage that the next 5 years would see identification of a large number of proteins and molecular biomarkers in urinary bladder cancer research. The main issue of translation from laboratory to clinical practice needs a focused approach. An increase in the application of panels of biomarkers rather than a single candidate is expected in the future, particularly in screening studies. It is anticipated that a combination of biomarkers and the existent surveillance protocol would reduce the number of follow-up cystoscopies, and thus the cost involved in the management of bladder cancer. Acknowledgements

NHS Grampian endowments funds for urinary proteomics research.

Expert commentary

Financial disclosure

At present, most of the urinary biomarkers are either being tested in conjunction with the standard care approach or they are still in translation phase of research. Whether or not they will replace cystoscopic examination is too early to say.

The authors have no relevant financial interests, including employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties related to this manuscript.


1113

Lam & Nabi

Key issues • Management of bladder cancer, especially superficial disease, requires life-long surveillance using invasive cystoscopic examination. • There is an urgent need to replace invasive follow-up procedures with noninvasive tests (urinary biomarkers). • -

Urinary biomarker research is continually evolving and a focus has shifted to following key issues: Laboratory validation of biomarkers found on initial screening using complementary techniques Statistical validation (sensitivity and specificity) in comparison with standard care approach Clinical effectiveness of approach using biomarkers either in combination with standard approach or as a screening test Cost–effectiveness of approaches relying on biomarker detection of bladder cancer

References 1

2

3

4

5

6

7

8

Jemal A, Siegel R, Ward E et al. Cancer statistics, 2006. CA Cancer. J. Clin. 56(2), 106–130 (2006). Holmang S, Hedelin H, Anderstrom C, Johansson SL. The relationship among multiple recurrences, progression and prognosis of patients with stages Ta and T1 transitional cell cancer of the bladder followed for at least 20 years. J. Urol. 153(6), 1823–1826; discussion 1826–1827 (1995). Botteman MF, Pashos CL, Redaelli A, Laskin B, Hauser R. The health economics of bladder cancer: a comprehensive review of the published literature. Pharmacoeconomics 21(18), 1315–1330 (2003). Halling KC, King W, Sokolova IA et al. A comparison of cytology and fluorescence in situ hybridization for the detection of urothelial carcinoma. J. Urol. 164(5), 1768–1775 (2000). Pfister C, Chautard D, Devonec M et al. Immunocyt test improves the diagnostic accuracy of urinary cytology: results of a French multicenter study. J. Urol. 169(3), 921–924 (2003). Grossman HB, Messing E, Soloway M et al. Detection of bladder cancer using a pointof-care proteomic assay. JAMA 293(7), 810–816 (2005). Nabi G, Greene DR, O’Donnell M. How important is urinary cytology in the diagnosis of urological malignancies? Eur. Urol. 43(6), 632–636 (2003). Glas AS, Roos D, Deutekom M et al. Tumor markers in the diagnosis of primary bladder cancer. A systematic review. J. Urol. 169(6), 1975–1982 (2003).

9

Brown FM. Urine cytology. It is still the gold standard for screening? Urol. Clin. North Am. 27, 25-37 (2000).

10

Duggan BJ, Gray SB, McKnight JJ, Watson CJ, Johnston SR, Williamson KE. Oligoclonality in bladder cancer:

1114

of cancer: overview of study designs and methods. Clin. Trials 3(1), 43–56 (2006).

the implication for molecular therapies. J. Urol. 171(1), 419–425 (2004). 11

Khan J, Saal LH, Bittner ML et al. Gene expression profiling in cancer using cDNA microarrays. Methods Mol. Med. 68, 205–222 (2002).

12

Velculescu VE, Zhang L, Vogelstein B, Kinzler KW. Serial analysis of gene expression. Science 270(5235), 484–487 (1995).

13

Marshall E. Getting the noise out of gene arrays. Science 306(5696), 630–631 (2004).

14

Wilkins MR, Sanchez JC, Gooley AA et al. Progress with proteome projects: why all proteins expressed by a genome should be identified and how to do it. Biotechnol. Genet. Eng. Rev. 13, 19–50 (1996).

15

Pandey A, Mann M. Proteomics to study genes and genomes. Nature 405(6788), 837–846 (2000).

16

Vlahou A, Schellhammer PF, Mendrinos S et al. Development of a novel proteomic approach for the detection of transitional cell carcinoma of the bladder in urine. Am. J. Pathol. 4, 1491–1502 (2001).

17

Lindon JC, Holmes E, Nicholson JK. Metabonomics techniques and applications to pharmaceutical research and development. Pharm. Res. 6, 1075–1088 (2006).

18

Pepe MS, Etzioni R, Feng Z et al. Phases of biomarker development for early detection of cancer. J. Natl. Cancer Inst. 93(14), 1054–1061 (2001).

19

Lokeshwar VB, Habuchi T, Grossman HB et al. Bladder tumor markers beyond cytology: International Consensus Panel on bladder tumor markers. Urology 66(6 Suppl. 1), 35–63 (2005).

20

Baker SG, Kramer BS, McIntosh M, Patterson BH, Shyr Y, Skates S. Evaluating markers for the early detection

21

van Rhijn BW, van der Poel HG, van der Kwast TH. Urine markers for bladder cancer surveillance: a systematic review. Eur. Urol. 47(6), 736–748 (2005).

22

Lokeshwar VB, Obek C, Pham HT et al. Urinary hyaluronic acid and hyaluronidase: markers for bladder cancer detection and evaluation of grade. J. Urol. 163(1), 348–356 (2000).

23

Konety BR. Molecular markers in bladder cancer: a critical appraisal. Urol. Oncol. 24(4), 326–337 (2006).

24

Steiner G, Schoenberg MP, Linn JF et al. Detection of bladder cancer recurrence by microsatellite analysis of urine. Nat. Med. 6, 621–624 (1997).

25

Lin Y, Miyamoto H, Fujinami K et al. Telomerase activity in human bladder cancer. Clin. Cancer Res. 2(6), 929–932 (1996).

26

Sanchini MA, Gunelli R, Nanni O et al. Relevance of urine telomerase in the diagnosis of bladder cancer. JAMA 294(16), 2052–2056 (2005).

27

Nielsen ME, Schaeffer EM, Veltri RW, Schoenberg MP, Getzenberg RH. Urinary markers in the detection of bladder cancer: what’s new? Curr. Opin. Urol. 16(5), 350–355 (2006).

28

Vriesema JL, Poucki MH, Kiemeney LA, Witjes JA. Patient opinion of urinary tests versus flexible urethrocystoscopy in follow-up examination for superficial bladder cancer: a utility analysis. Urology 56(5), 793–797 (2000).

29

Stoeber K, Swinn R, Prevost AT et al. Diagnosis of genito-urinary tract cancer by detection of minichromosome maintenance 5 protein in urine sediments. J. Natl Cancer Inst. 94(14), 1071–1079 (2002).



30

Black PC, Brown GA, Dinney CP. Molecular markers of urothelial cancer and their use in the monitoring of superficial urothelial cancer. J. Clin. Oncol. 24(35), 5528–5535 (2006).

Website 101

National Cancer Institute. Early Detection Research Network http://edrn.nci.nih.gov


Affiliations •

Thomas Lam University of Aberdeen, Academic Urology Unit First floor, Health Sciences Building, Aberdeen AB25 2ZD,UK Tel.: +44 122 455 4877 Fax: +44 122 455 4580 [email protected]

•

Ghulam Nabi University of Aberdeen, Academic Urology Unit, First floor, Health Sciences Building, Aberdeen, AB25 2ZD,UK Tel.: +44 122 455 4877 Fax: +44 122 455 4580 [email protected]

1115