SEARCH

SEARCH BY CITATION

Keywords:

  • bladder cancer;
  • molecular markers;
  • urothelial cell carcinoma

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Many markers for the detection of bladder cancers have been tested and almost all urinary markers reported are better than cytology with regard to sensitivity, but they score lower in specificity. Currently molecular and genetic changes play an important role in the discovery of new molecular markers for detection, prognostication and surveillance. The purpose of this review is to highlight the most important urinary molecular biomarker developments that have been studied and reported recently. In the current review we have summarized the most recent and relevant published reports on molecular urinary markers. The results of this review show that the first generation of urinary markers did not add much to urinary cytology. The current generation of markers is better, but additional clinical trials are needed. Our knowledge of molecular pathways in bladder cancer is growing and new methods of marker development emerge, but the perfect marker is still to be found. Currently, there are not clinically usable molecular markers that can guide us in diagnosis or surveillance, nor guide us in lowering the frequency of urethrocystoscopy in bladder cancer.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Bladder cancer is the second most common malignancy affecting the urinary system. Approximately 90% of urothelial tumors are urothelial cancers (UC). The remaining tumors are squamous cell cancers or adenocarcinomas. The spectrum of bladder cancer includes low- and high-risk non-muscle-invasive, muscle-invasive and metastatic disease. Each group has its own specific typical behavior, prognosis and treatment. Diagnosis is based on cystoscopy and urinary cytology. Although considered the gold standard, urethrocystoscopy can miss certain lesions, in particular small areas of carcinoma in situ (CIS). Moreover, urethrocystoscopy is invasive for the patient and it is expensive. Cytology, the second gold standard, has a median sensitivity of only 35%, and a median specificity of 94%,1 although it is useful for detecting high-grade tumors and CIS. However, because of its low sensitivity in low-grade cancers it has limited clinical relevance. The limitations of cytology and the invasiveness of urethrocystoscopy have generated interest in other non-invasive tools. Our increasing understanding of molecular processes involved in the development of cancer has resulted in the development of new methods of diagnosing, surveying and prognosticating cancers in general, and UC in particular. The molecular and genetic changes in UC of the bladder can be broadly classified into three interrelated processes: (i) chromosomal alteration, triggering the initial carcinogenic event; (ii) tumor proliferation, caused by loss of cell-cycle regulation and derangements in normal apoptotic turnover; and (iii) metastasis, in which the initial tumor spreads and bring into play processes such as angiogenesis and loss of cell adhesions.2 Obviously, in different stages of the disease different changes in the physiological process, caused by environmental factors, stress, injury or disease, are apparent. As a result, different biomarkers will be applicable in different phases of cancer development, such as detection, progression and metastasis.

Developing and using molecular biomarkers employs techniques of modern molecular biology that investigate the three basic pathways of gene expression: the gene itself, the messenger ribonucleic acid (mRNA) that it produces, and the protein that is coded for by the mRNA.

Molecular assays

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Not very long ago we diagnosed the presence of cancer with histopathology. Nowadays new technologies, like high-throughput genomic and proteomic assays, help us understand molecular pathways in bladder cancer.

Genomics

Gene microarray technology rests on the ability to deposit many (tens of thousands) different DNA sequences on a small surface, usually a glass slide (often referred to as a ‘chip’). These chips can be used to measure the mRNA expression levels for tens of thousands of genes from a tissue sample by hybridizing fluorescently labeled complementary DNA from that tissue to the chip.3 This technology is a great improvement compared to earlier techniques for measuring gene expression (northern blot, reverse transcriptase–polymerase chain reaction [RT–PCR]). These methods are only capable of studying a few genes per experiment.

Tissue microarray (TMA) makes it possible to study very large numbers of tumors at once, either at the DNA, RNA or protein level. This analysis may be based on fluorescence in situ hybridization (FISH) for gene expression or immunohistochemistry for protein over expression.4

Gene microarray and TMA are complementary in the features that can be examined. While gene micro arrays are used to test the level of expression for many thousands of genes in relatively few samples, TMA are used to test the staining pattern of relatively few markers on a large number of samples.3 Both microarray technologies may be combined to rapidly validate gene targets.

Microarray can, as mentioned above, provide a powerful tool for characterizing gene expression on a genome scale. Many studies using micro arrays have been performed concerning various types of cancer to find cancer-type-specific gene expression signatures. The diversity of these signatures makes it difficult to distinguish the genes that play a crucial role in oncogenic processes from those that are spuriously differentially expressed and therefore irrelevant to the oncogenic process.5 Appanna6 studied a group of 16 patients with bladder augmentations and, using comparative genomic hybridization (CGH) or chromosomal microarray analysis (CMA), found genetically unstable regions. They found that urothelium adjacent to the bladder and/or bowel anastomosis in these patients exhibited a greater degree of genetic instability than that distant from the anastomosis. Further studies are needed but this study makes clear that this field of investigation opens new possibilities for screening those who are at risk.

Michiels et al.7 reanalyzed data from the seven largest published studies that have attempted to predict prognosis of cancer patients on the basis of DNA microarray analysis between January 1995 and April 2003. They found that the list of genes identified as predictors of prognosis was highly unstable and that molecular signatures strongly depended on the selection of patients. They stated that the prognostic value of microarray results should be considered with caution and advocated the use of validation by repeated random sampling. Wang et al.8 stated that the problem is the absence of a ‘gold standard’ data set that allows an evaluation of different microarray platforms based on a common ‘ground truth’. One strategy is using a well accepted reference data set generated by a reliable independent technology, such as real-time PCR for gene expression measurements. For bladder cancer, such a validation study was performed by Schulz et al.9 Their aim was to validate a panel of 26 genes, identified from a large complementary DNA microarray analysis of bladder tumors that discriminated between early- and late-recurring patients with non-muscle-invasive bladder tumours.10 They used real-time quantitative PCR, but were unable to confirm the microarray gene expression pattern. They concluded that accurate identification of patients with short and prolonged recurrence free periods using molecular markers remains a challenge and may depend on the right combination of several predictive genes. Also, the results of microarray in the discovery of such genes need careful validation.

In conclusion, gene microarray technology makes it possible to study many genes simultaneously; it can help us understand molecular processes in cancer and bears great potential for future molecular marker development. It is a relatively new field of science and validation of discovered genes remains necessary.

Proteomics

In proteomics the structure and function of proteins is studied. This study plays an important role in the search for molecular cancer biomarkers. Innumerable proteins can be found in human urine; the proteome reflects the cellular state or the external conditions encountered by a cell, and proteome analysis can be viewed as a genome-wide assay to differentiate and study cellular states and to determine the molecular mechanisms that control them.11 Changes in excretion rates of specific proteins/peptides can have predictive value in the early diagnosis and surveillance of bladder cancer. The wide availability of urine-samples and the noninvasive nature of urinary proteomics mean that this analysis is very appropriate for clinical research.

The original proteomic approach is two-dimensional electrophoresis (2DE) followed by mass-spectrometry (MS).12 Using 2DE the proteins presented in a sample are separated in two dimensions: in the first separation proteins are grouped according to isoelectric points, and the proteins with identical isoelectric points are than separated in the second dimension, protein mass. This technique is also called gel-electrophoresis because it makes use of a gel and the above mentioned procedure will lead to a spot-like gel pattern. These spots can be examined with MS. This technique is used to simply measure the molecular mass of the visualized proteins. The end result is an extensive list of the proteins that are present in a specific sample.

Many studies have been performed to identify protein that can be used in clinical settings. Known protein diagnostic biomarkers are nuclear matrix protein 22, bladder tumor antigen and BLCA-4 and will be described in more detail.

A newer technology uses ProteinChip technology. Using ProteinChip array technology, the proteome is simplified into sub-proteomes by retentate chromatography. This procedure is followed by a detection method using surface-enhanced laser desorption/ionization (SELDI) time-of-flight-mass spectrometry (TOFMS).13 Retentate chromatography is performed on ProteinChip arrays with varying chromatographic properties. By utilizing arrays with different surface chemistries in series, a complex mixture of proteins, from cells or body fluids, can be resolved into subsets of proteins with common properties. Then the arrays are washed to remove the weakly bound protein and a matrix is added and allowed to crystallize after which the arrays are read in a ProteinChip reader.13

In a training set of 30 archived urine samples from bladder cancer patients and 30 urinary samples from healthy volunteers proteins were analyzed using ProteinChip technology and computer-based data mining. Bladder cancer was segregated from control with a sensitivity of 80% and specificities of 90% to 97%.14 Munro et al.15 obtained duplicate proteomic profiles from 227 subjects (118 UC, 77 healthy controls and 32 controls with benign urological conditions). They used linear mixed-effect models to identify peaks that are differentially expressed between models and within UC subgroups. From this group 130 profiles were randomly selected in an initial test set (n = 43) and an independent validation set (n = 43). UC was predicted with 71.7% sensitivity and 62.5% specificity in the initial set and with 78% sensitivity and 65.0% specificity in the validation set.

In conclusion, proteomics is very promising in marker development. Proteins reflect the cellular state and new technologies make it possible to study the numerous proteins that can be found in human urine. It's very likely that this will lead to the discovery of new molecular markers.

Molecular markers in urothelial cancers

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Fluorescence in situ hybridization

As in other forms of cancer, bladder cancer also shows chromosomal anomalies. With FISH techniques, chromosomal anomalies can be detected in exfoliated bladder cells. A currently commercially available test, the Urovysion test, has probes for chromosome 3, 7, 17 and a locus specific probe for 9p21.

In various case–control studies, as recently summarized by Lokeshwar16 the sensitivity of FISH varies between 69% and 87%. All studies reported a low sensitivity of FISH to detect low-grade (36%–57%) and low-stage (62%–65%) tumors, but FISH has high sensitivity to detect high-grade and high-stage tumors (83%–97%). The detection of CIS is close to 100%. The specificity of FISH is high (89%–96%) and is comparable to cytology. The limited performance of FISH in low-grade or low-stage tumors is not consistent. Jones17 also recently reviewed published reports on the role of FISH in bladder cancer surveillance. He concluded that FISH outperformed conventional cytology across all stages and grades in all published reports. Notably, cytology detected only 67% of CIS versus 100% detection by FISH. Marin-Aguilera et al.18 found in their study higher overall sensitivity for FISH versus cytology (70.3% vs 35.1%). In this study the significant difference was maintained when non-muscle-invasive UC detection was broken down into low-grade and high-grade tumors. In contrast with these two studies, Moonen19 found no improvement of FISH over cytology in the diagnosis of recurrent non-muscle-invasive bladder cancer in a study including 64 patients with biopsy-proven UC. Sensitivity and specificity were, respectively, 39.1% and 89.7% for FISH, and 40.6% and 89.7% for cytology.

A potential advantage of FISH is its ability to detect occult diseases not visible on urethrocystoscopy. Many authors note that a false positive FISH test can predict future recurrence within 3 to 12 months in 41% to 89% of patients.20–22 Veeramachaneni et al.23 concluded in a cohort study that a positive FISH test may indicate frank neoplastic urothelial transformation, or it may merely be an indicator of unstable urothelium. Specificity of FISH, therefore, may be underestimated because of this phenomenon and explains why FISH performs different in different patient populations, for example less in surveillance compared to detection of primary tumors. Another advantage of FISH is that it is unaffected by bacillus Calmette–Guérin (BCG) therapy and can therefore also be used for surveillance in patients who have been treated with intravesical BCG.24 Mengual et al.25 used FISH to determine the response of patients with high-risk superficial bladder tumors to BCG therapy. Bladder washing specimens were retrieved from 65 patients and after BCG treatment they were analyzed using FISH. Pre-BCG, 85% had a positive FISH result and after BCG treatment, 45% had a positive and 35% had a negative FISH result. They found that patients with a positive post-BCG FISH status had a 2.7-fold greater risk of tumor recurrence than patients with a negative post-BCG status. Thus FISH also appears to be useful for the surveillance of patients treated with BCG.

A disadvantage is that this test is labor intensive and there is a clear learning curve before this test can be used reliably.

In conclusion, because of the workload and the high costs of the FISH test its clinical use is still low. However, most authors agree that FISH is better than cytology, although the test might have limited sensitivity to detect low-grade tumors. The fairly high false positive rate is explained by some to reflect the potential of the FISH test to predict future recurrences.

Microsatellite analysis

Microsatellites are highly polymorphic short tandem DNA repeats found in the human genome. Two types of microsatellite alterations can be found in many cancers: loss of heterozygosity (LOH), an allelic deletion and somatic alteration of microsatellite repeat length.26 In bladder cancer, most mutations are in the form of LOH.26,27 Microsatellite alterations in exfoliated urine are detected by a PCR using DNA primers for a panel of known microsatellite markers.

In various studies, overall sensitivity and specificity of microsatellite analysis (MSA) ranged from 72% to 97% and 80% to 100%, respectively.16 In contrast to conventional cytology it appears that MSA has the ability to detect low-grade and low-stage disease as accurately as high-grade and high-stage disease.28 Frigerio et al.29 found that the combined use of cytology and LOH-analysis had high sensitivity for identifying primary tumors and had the ability to detect almost all recurrent diseases in voided urine. They found the sensitivity for grade 1–2 tumors was 72% and that for grade 3 tumors was 96%.

Recently van der Aa30 assessed the feasibility and clinical utility of MSA on voided-urine samples in a routine setting to detect or predict bladder cancer recurrences. They evaluated 228 patients who participated in a prospective multi-centre study in 10 hospitals. They were randomized after resection of non-muscle-invasive UC into two trial arms whose cystoscopy and synchronous MSA data were both available in 1012 follow-up visits. DNA from urine was amplified by PCR using primers for 20 polymorphic microsatellite markers localized on 10 chromosomes. Interestingly, 19% of samples could not be analyzed for reasons of low DNA quality. Most LOH were found with the marker for chromosome 9. They found a relatively low sensitivity of 58% and a specificity of 73%. A potential limitation of their study is that their main analysis is based on the combined data from the two arms. Separate analysis of the patients in the MSA arm showed an improved sensitivity (70%) of the MSA test. However, MSA on voided-urine samples is currently not sufficiently sensitive to recommend implementation in routine clinical practice.

In conclusion, MSA has good overall sensitivity and specificity, but this test is complex and expensive and therefore not used in daily clinical practice.

Immunocyt

Immunocytology is based on the visualization of tumor-associated antigens in urothelial carcinoma cells using monoclonal antibodies. Three fluorescently marked antibodies label two mucin-like proteins and a high molecular weight form of carcinoembryonic antigen. After this process the cells are examined under a fluorescence microscope. Sensitivity varies between 38.5% and 100%.31–34 Immunocyt shows specificity between 73% and 84.2%.31–33 A prospective study in which 942 patients were enrolled showed 298 patients with histopathologically proven UC. The results were encouraging: sensitivity for grade 1 tumors was 79.3%, 84.1% for grade 2 and 92.1% for grade 3. Specificity was 72.5%.35 Schmitz-Dräger et al.34 recently found high sensitivity and good specificity in a population of 189 patients with microhematuria. They found bladder cancer in eight patients and only one tumor of low malignant potential was missed. They concluded that the high sensitivity and good specificity in a population with a low disease prevalence could have prevented 154 costly and invasive diagnostic procedures. A prospective study, also by Schmitz-Dräger et al.,36 in which the role of immunocytology in patients with gross hematuria was assessed, showed that the combination of cystoscopy and immunocytology gave 100% sensitivity, while combining cystoscopy and cytology marginally improved the sensitivity of cystoscopy alone. In this study they included 61 consecutive patients with a first episode of painless gross hematuria, but no previous UC. Sensitivity for cystoscopy, immunocytology and cytology were 76%, 88%, and 47% and specificity was 100%, 77%, and 95%, respectively. The most important observation was that the combination of cystoscopy and immunocytology gave 100% sensitivity.

In conclusion, sensitivity of immunocyt is good, but it fails in comparison to conventional cytology with regard to specificity. Another aspect is the high inter-observer variability and the need for constant quality control.

Telomerase

Telomeres are repetitious sequences at the end of chromosomes that protect genetic stability during DNA replication. There is loss of telomeres during each cell division, which causes chromosomal instability and cellular senescence. Bladder cancer cells express telomerase, an enzyme that regenerates telomeres at the end of each DNA replication and therefore sets the cellular clock to immortality. Three major subunits comprising the human telomerase complex have been identified: a ribosomal RNA component (human telomerase [hTR]) that serves as a template for telomere repeat synthesis; a protein component (human telomerase reverse transcriptase [hTERT]) responsible for enzymatic activity of telomerase enzyme; and finally telomerase enzyme associated with proteins with an unclear function.37 Determination of telomerase activity is a PCR-based technology and must be performed in specialized laboratories. Eissa et al.38 evaluated three methods of telomerase detection for diagnostic accuracy. They included 200 patients with bladder cancer, 85 with benign bladder lesions and 30 healthy individuals. From voided urine they analyzed relative telomerase activity by telomeric repeat amplification protocol (TRAP assay), human telomerase RNA (hTR-mRNA) by RT–PCR and hTERT by RT–PCR. Overall the sensitivity of human telomerase reverse transcriptase for detecting bladder cancer was the highest compared to that of human telomerase RNA, relative telomerase activity and urine cytology (96%, 92%, 75% and 75%, respectively). Specificity was 96%, 89%, 92% and 94%, respectively. hTERT followed by hTR provided sensitivity of 100% and specificity of 85% and thus an improvement in diagnosis of bladder cancer. Overall sensitivity and specificity of the telomerase assay, as reported by Lokeshaw16 were 70%–100%, and 60%–70%, respectively. In a systematic review Glas et al.39 showed that telomerase had the best sensitivity (75%) compared to the other markers, including cytology. Specificity, however, was lower than cytology. A more recently conducted case control study on 218 men showed an overall sensitivity of 90% and specificity of 88%, which increased to 94% for individuals younger than 75 years of age.40 The same phenomenon was noted by Bravaccini:41 in a case–control study conducted in 212 women, sensitivity was 87% and specificity 66%. A breakdown analysis as a function of age showed a higher assay accuracy in women younger than 75 years (sensitivity 91% and specificity 69%) compared to older women (sensitivity 64% and specificity 59%). The authors noted that a possible explanation for this effect is the higher number of viable and telomerase-positive nonurothelial cells, such as epithelial cells from the lower genital tract or inflammatory elements in urine of the elderly. Bian et al.42 studied a combined assay of CYFRA21-1, telomerase and vascular endothelial growth factor in the detection of bladder cancer. They collected 100 midstream urine specimens with known bladder cancer and each sample was aliquoted for the various assay. The found a sensitivity for telomerase of 71% and specificity of 84%.

In conclusion, it seems that telomerase has good sensitivity but lacks sufficient specificity, however new methods of telomerase detection improve specificity. Moreover, test results can be influenced by inflammation and age. These disadvantages make it a suboptimal test for detection of bladder cancer.

Hyaluronic acid and hyaluronidase

Hyaluronic acid (HA) is a glycosaminoglycan and a normal component of tissue matrices and body fluids. In tumor tissues, elevated HA is mostly localized to tumor stroma. In bladder carcinoma HA is found in tumor cells and HA levels have shown to be elevated in urinary samples of bladder cancer patients.43 The concentration of HA is also associated with tumor metastases.44 Hyaluronidase (HA-ase) is an enzyme that cleaves HA into fragments. Three HA-ase genes have been identified and bladder tumor derived HA-ase was shown to be the HYAL1 type.43 HA-ase levels are elevated in bladder tumor tissue, and an increase is correlated with tumor grade.45 In a study by Lokeshwar et al.43 it was shown that blocking HYAL1 expression in a bladder cancer cell line results in a 4-fold decrease in cell growth rate, suggesting that HYAL1 expression by tumor cells is required for cell proliferation. They also mentioned that HYAL1 plays a role in promoting the invasive potential of bladder tumor cells and is not elevated in low-grade tumors. Both the HA-test and HA-ase-test are enzyme linked immunosorbent assays (ELISA). The HA test detects bladder cancer regardless of tumor grade and, as mentioned above, the HA-ase test preferentially detects grade 2 and 3 bladder tumours.46 Sensitivity of the combined HA-HA-ase test varies between 83% and 94%.16,47–50 The overall specificity varies between 77% and 93.4%.46–50 A study by Passerotti et al.51 in which urine samples were taken from 83 patients (22 controls and 61 diagnosed positive for bladder cancer) showed sensitivity of 92.9% and specificity of 83%. In a few comparative studies HA-HA-ase was shown to be better than other markers, including cytology.47–50

In conclusion: HA-HA-ase is a very promising marker that deserves further study. The test has high sensitivity to detect both low- and high-grade/stage tumors.

Nuclear matrix protein 22

Nuclear matrix protein 22 (NMP22) is a nuclear matrix protein and is an important regulator of mitosis. In tumor cells the nuclear mitotic apparatus is elevated and NMP22 is released from cells in detectable levels. The first NMP22 test was a quantitative ELISA test. The newer NMP22 ‘BladderChek’ is a point of care assay using monoclonal antibodies in a lateral flow strip, detecting the nuclear matrix protein NMP22 with a cut-off value of 10 U/mL. Grossman et al.52 investigated the capability of this test in detecting malignancy in 1331 patients with risk factors of bladder cancer. They found sensitivity of 55.7% and specificity of 85% for NMP22 as compared to 15.8% and 99.2% for cytology. In a subsequent study the same authors examined whether NMP22 could improve detection of recurrences in 668 patients. Sensitivity of NMP22 in this study was 49.5% and specificity was 87.3%. NMP22 detected eight malignancies that were not detected by urethrocystoscopy. A combination of NMP22 and urethrocystoscopy identified 99.0% of all malignancies versus 91.3% with urethrocystoscopy alone. Voided cytology did not significantly increase the sensitivity of urethrocystoscopy.53 Taking these studies into account, it seems that NMP22 performs less in surveillance compared to primary detection of bladder cancer, but still has a better sensitivity than cytology. In a recent study, NMP22 was compared with photodynamic diagnosis (PDD) as the gold standard.54 They collected urine samples from 100 patients with suspicion of bladder cancer. Then they performed the NMP22 BladderChek test and voided urine cytology. Subsequently the patients underwent PDD. The authors found a sensitivity of 65% and a specificity of 40% for NMP22. Voided cytology scored 44% and 78%, respectively. Washed cytology, however, had a sensitivity of 75% and a specificity of 62%. In contrast, PDD scored a sensitivity of 93% and a specificity of 43%. The value of the NMP22 test was limited by its low sensitivity, presumably due to frequent positive reactions in benign conditions. The authors concluded that the NMP22 test, validated by PDD cannot be recommended for screening or surveillance in daily routine use. Another point of debate is the cut-off value for the NMP22 test. Shariat et al.55 assessed the variability in the diagnostic performance of NMP22 for detecting recurrence and progression in a population of 2871 patients. With the manufacturer's recommended cut-off of 10 U/mL, 57% of cases were detected, with a 19% false-positive rate. They found a substantial degree of heterogeneity in the diagnostic performance of NMP22 applied to populations from different institutions, and stated that there is no clearly defined NMP22 cut-off at which to recommend urethrocystoscopy.

In conclusion, the current NMP22 point of care test is easy to perform, with a sensitivity better than cytology and a reasonable specificity. However, its sensitivity seems to be hampered by benign conditions.

BLCA-4

There are six nuclear matrix proteins identified that are specifically expressed in bladder cancer, and these proteins were termed as the BLCA proteins. One of them is BLCA-4. Over expression of BLCA-4 seems to increase the growth rate in cells and also causes cells to express a more tumorigenic phenotype.56 BLCA-4 is analyzed with ELISA and has a reported sensitivity of 89%–96.4% and its specificity ranges between 95% and 100%.57,58 A large multicentre trial is currently being performed.

BLCA-1, another nuclear matrix protein expressed in bladder cancer, shows a sensitivity of 80% and a specificity of 87% in a study involving 25 patients with bladder cancer and 46 controls. A limitation of this study was the small number of low-grade tumours.59

In conclusion, BLCA-4 seems to have good sensitivity and specificity for detecting bladder cancer, but a larger trial is needed to confirm this. There is as yet not much evidence about the performance of BLCA-1 in detecting bladder cancer.

Other nuclear matrix proteins

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Further proteomic analysis is done with regard to nuclear matrix (NM) proteins Barboro et al.60 tried to identify NM proteins from bladder tumor tissue specimens that might serve as prognostic markers and focused on molecular analysis of NM changes associated with tumor development in muscle invasive UC. They analyzed 21 samples of patients who had undergone radical cystectomy and nine samples of adjacent, non-tumoral tissue and healthy urothelium were collected from three unaffected individuals. They used 2DE, and selected spots were characterized by liquid chromatography coupled to tandem MS and western blot. Thirty proteins were differentially expressed by bladder tumor cells; among these, 19 proteins were detected in bladder tumoral tissues but not in normal and non-tumoral tissues and seven proteins correlated with tumor stage. They found that p54nrb expression was strongly correlated with the patient's mortality. P54nrb is a nuclear RNA-and DNA-binding protein whose specific functions are not known, but the authors found a strong relation between expression of this protein and vascular invasion. Because these findings are very premature, more investigation is necessary.

Cytokeratins

Cytokeratins are intermediate filaments, their main function is to enable cells to withstand mechanical stress. In humans 20 different cytokeratin isotypes have been identified. Cytokeratins 8, 18, 19 and 20 have been associated with bladder cancer.61

The urinary bladder cancer (UBC) test detects cytokeratins 8 and 18 fragments in urine. Sensitivity of the UBC-test varies from 35% to 79% and depends on tumor grade and stage 2, but UBC-tests were inferior to voided cytology in test quality.16,62

CYFRA 21-1 is a soluble fragment of cytokeratin 19 and is analyzed with ELISA. It is measurable in serum and urine. Abnormal serum levels of CYFRA 21-1 in patients with bladder cancer were only seen in patients with metastatic disease.63 Abnormal CYFRA 21-1 levels also showed a significantly worse overall median survival, and correlated with response to systemic treatment.64 In an early study Pariente et al.65 found an optimal cut-off value of 4 ng/mL for urinary CYFRA 21-1 for diagnostic performance of urinary CYFRA 21-1. In a prospective study urine samples of 325 patients were examined, including 107 patients under surveillance after transurethral resection of bladder cancer. This study found an optimal cut-off concentration for the detection of primary bladder tumors of 4.9 ng/mL. This cut-off resulted in a sensitivity of 79.3% and a specificity of 84%.66 These authors also found increased concentrations of CYFRA 21-1 after instillation with BCG, even years after treatment. A recent study by Fernandez-Gomez67 showed sensitivity of 43% and specificity of 68% at a cut-off value of 4 ng/mL. Lowering the cut-off point to 1.5 ng/mL increased sensitivity to 73.8% but decreased specificity to 41%. Specificity increased excluding all patients treated with pelvic radiotherapy, with urinary tract infections (UTI) or urethral catheterization and intravesical instillation within the previous 3 months. With a cut-off value of 4 ng/mL sensitivity increased to 80.2%. In a study by Bian et al.,42 cut-offs for the CYFRA21-1 assay were 3.5 ng/mL and these authors found a sensitivity of 74% and a specificity of 78%. In contrast, this study showed a sensitivity and specificity of cytology of 38% and 92%, respectively.

CK20

Cytokeratin 20 is expressed in the umbrella cells of normal urothelium. These umbrella cells are highly differentiated superficial cells with two unique characteristics: a sub-apical network of cytokeratins and membrane protein plaques (which has been nominated as playing a role in preventing urothelial rupturing during bladder distension); and they contribute to the permeability barrier of the urothelium.68 Van Oers et al.69 found that 65% of pTa tumors revealed an abnormal CK20 pattern, which implies that deregulation of CK20 is an early event in the development of UC. But in pTaG1 tumors, FGFR3 mutations were more frequent (82%) than an abnormal CK20 pattern (59%). The presence of a normal CK20 pattern and of a mutation in the FGFR3 gene correlated with the group of pTa G1-2 tumors, while FGFR3 mutant tumors with an abnormal CK20 are less differentiated, and some of higher stage (18%). They concluded that the combination of FGFR3 and CK20 could be an excellent prognostic marker. In a small study Bhatia et al.70 looked at the utility of CK20 immunostaining in identifying malignant cells in urine cytology smears. Fourteen cases, each with an unequivocal diagnosis of UC, were collected. Fourteen cases of benign urinary cytology and five cases with a diagnosis of atypical cells were also subjected to immunohistochemistry. Twelve cases in the UC group stained positive with CK20, indicating high sensitivity (86%). All cases with benign cytology were negative, indicating a high specificity (100%). The five atypical cytology cases were also positive for CK20. Pu et al.71 evaluated the expression of survivin, CK20 and MUC7 mRNA in voided urine of patients with bladder cancer. Voided urine of 153 patients and 20 healthy volunteers were evaluated by RT–PCR. The three markers were evaluated independently or in combinations. They found a sensitivity for CK20 of 82.6% and a specificity of 97.4%. The combined sensitivity of the three markers was higher than that of the markers alone.

In conclusion, CYFRA 21-1 shows a disappointing performance in low-stage bladder cancer, and CYFRA21-1 levels are strongly influenced by benign urological diseases and intravesical instillations. CK20 seems to be a good marker for prognostication in which normal CK20 is associated with low-grade tumors.

BTA-TRAK and BTA-stat

BTA-TRAK and BTA-stat are both versions of the bladder tumor antigen assay that measure complement factor H-related protein in urine. BTA-stat is an immunoassay that can be performed ‘on bench’ within several minutes. BTA-TRAK is a quantitative test that is performed in a laboratory. The published reports recently reviewed by van Rhijn1 showed a sensitivity that was slightly higher than cytology, but the specificity was much lower. BTA-stat had a median sensitivity of 70% (range 24% to 89%) and median specificity of 75% (range 52% to 93%). For BTA-TRAK, median sensitivity was 69% (range 57% to 79%) and median specificity was 65% (range 48% to 95%). These tests can also be false positive in patients with inflammation, infection or hematuria.1

In conclusion, the usage of BTA-TRAK and BTA-stat is limited because of its low specificity and because the false positive test results in patients with benign genitourinary conditions.

Glycoproteins

In a pilot study Kreunin et al.72 analyzed one specific fraction of the urinary proteome, that of glycosylated proteins. Alterations in protein glycosylation have been shown to correlate with numerous diseases, including cancer. The most well-known glycoprotein biomarker is the prostate-specific antigen (PSA) in prostate cancer. The authors identified 186 proteins from eight urine samples and found that the presence of several glycoproteins was associated with UC.

But the role of glycoproteins as markers for UC hampers further study to evaluate their true potential in complex biological samples and this facilitates the identification of potential biomarkers of UC in noninvasively obtained human urine.

Survivin

Survivin is a member of the family of proteins that regulates cell death, the so-called inhibitor of apoptosis family. Its over expression inhibits extrinsic and intrinsic pathways of apoptosis.73 It is expressed in the G2/M phase of the cell cycle and has been shown to be involved in the regulation of chromosome alignment and segregation.74 Survivin is expressed during fetal development but is not expressed in terminally differentiated adult tissues,75 however it is one of the most commonly over-expressed genes in cancer.76 In bladder cancer there is expression of survivin in urine, and its expression is associated with disease recurrence, stage, progression and mortality.77 RT–PCR provides a diagnostic tool to detect survivin mRNA in urine. In recent published reports, sensitivity and specificity from 64% to 94% and 93% to 100%, respectively, have been noted.73,78,79 Schultz et al. looked at the mRNA expression of 23 genes in 44 primary pTa tumours,80 and survivin mRNA expression helped to distinguish between long or short recurrence free intervals in patients with primary Ta UC. Survivin identified 71.4% and 69.6% of the patients with long or short recurrence-free periods, respectively. Chen et al.74 studied the use of survivin and Ki-67 as potential markers for grading non-muscle-invasive UC They evaluated the interobserver variability of grading low-grade and high-grade UC using the 2004 World Health Organization system. In this study 51 bladder biopsies were graded blindly by five experienced pathologists. The protein and mRNA expression profiles of survivin and Ki67 were analyzed using immunohistochemistry and RT–PCR. Survivin outperformed Ki67 in separating the high-grade group from the low-grade group and showed a significantly higher predictive accuracy for high-grade recurrence than the histological grade.

In conclusion, survivin is a very promising marker with good sensitivity and very good specificity and also deserves further study. Survivin seems predictive for recurrence and can be helpful in preventing unnecessary urethro-cystoscopies.

Cytokines

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Tumor necrosis factor

Tumor necrosis factor (TNF)-α and TNF-β are structurally related cytokines that are secreted by macrophages and lymphocytes, respectively.81,82 The TNF-genes contain a relatively large number of polymorphisms, possibly associated with a susceptibility to bladder cancer and prognosis. Nonomura et al.83 looked at a possible correlation between polymorphism in the TNF-β and TNF-α gene and clinical features of bladder cancer in a group of 141 Japanese patients with bladder cancer and 173 Japanese controls with benign disease. Genotyping was done through PCR of blood samples. They could not find a correlation between TNF-α polymorphisms and clinicopathological parameters. In the TNF-β group, however, there was a significant difference in the genotype distribution between the bladder cancer patients and the controls. Patients with the β1/2 genotype showed a 1.71-fold increased risk of bladder cancer compared with the β2/2 genotype. In contrast, Kim et al.84 performed PCR of blood samples of 153 patients with primary bladder cancer and 153 control subjects in an earlier study, and found a significant increase of the TNF-α genotype and cancer stage. They concluded that these data suggest that these genetic polymorphisms may be useful as prognostic markers for bladder cancer in the clinical setting. A possible explanation for these differences are given by Nonomura et al.83: the frequency of TNF-α polymorphisms appears to differ among ethnic groups.

In conclusion, TNF can have prognostic value, but the polymorphisms can differ between ethnic groups. More clinical studies are necessary to define the place of TNF as a marker of progression.

Growth factors

Fibroblast growth factors receptor 3

Fibroblast growth factors (FGF) represent a large family of polypeptides that are potent regulators of cell proliferation, migration and differentiation. FGF receptors (FGFR) belong to the family of tyrosine kinases, an enzyme that plays an important role in activation of several different intracellular signaling pathways. In a study by Rieger et al.85 they took urine sediment DNA samples from 192 patients, of whom 72 had undergone transurethral resection (TURBT) of mainly Ta lesions and 120 had undergone cystectomy. The patients in the cystectomy group had more advanced tumors than those in the TURBT group. They found that 67% of patients in the TURBT group and 28% of patients in the cystectomy group displayed FGFR3 mutations. In 122 cases comparative analysis of cytology was performed and FGFR3 analysis identified change in 68% of urine sediment DNA, whereas cytology recorded the presence of tumor cells in 32% of the DNA samples. But more importantly it seems that FGFR3 is expressed in low-grade and low-stage disease. Van Rhijn et al.86 performed a multicenter study, which included 286 patients with primary urothelial cell carcinoma, and found that FGFR3 mutations occur in 60% of cases of primary urothelial cell carcinoma. Eighty-eight percent of grade 1 tumors had an FGFR3 mutation while in grade 3 tumors this rate was only 16%. So the presence of an FGFR3 mutation was linked to a favorable disease course. This was confirmed by Gomez et al.,87 who used a micro array to evaluate FGFR3 mRNA expression in urinary carcinomas at different stages. They also showed a marked over-expression of both FGFR3 mRNA and protein in UC with a greater expression that was observed in stages pTa and pT1. Van Oers et al.88 examined four markers (Ki-67, TP53, CK20, and FGFR3). In this study an FGFR3 mutation analysis was performed in a group of 255 unselected patients with a primary UC. Mutations in the FGFR3 gene were detected in 47% of 208 bladder tumors, of which 61% occurred in pTa and pT1 tumors. FGFR3 mutations were predominantly present in tumors of stage pTa (73%) and grade 1 (82%). Almost all FGFR3 mutations occurred in tumors without adjacent carcinoma in situ, only 1% concurrence. Because of this observation they state that FGFR3 mutations and carcinoma in situ must be considered as exclusive events and FGFR3 mutations are associated with favorable histopathological characteristics. The authors concluded with regard to FGFR3 that it is a useful marker to identify superficial tumors with low malignant potential but that grade and stage are still better predictors.

In conclusion, FGFR3 plays an important role in the development of low-grade and non-muscle-invasive UC. Used as a molecular marker it is diagnostic for low-grade/low-stage and FGFR3 mutations seem to be associated with favorable tumor characteristics. This development gives a lot of information about molecular pathways that occur in the development of UC.

VEGF

Angiogenesis or the development of new blood vessels from the surrounding vasculature is essential for the growth and progression of solid tumors. Vascular endothelial growth factor (VEGF), a positive regulator of angiogenesis, plays a pivotal role in tumor angiogenesis and shows a high expression in almost all known tumors including UCs.89 Yang et al.90 studied the expression of VEGF in UC. Tissue samples from 161 patients with UC were examined with immunohistochemical staining for the expression of the VEGF gene. The expression rate was compared to 32 normal bladder mucosal samples obtained from transurethral surgery from non-cancer patients. The VEGF gene was expressed in none of the 32 normal bladder mucosal samples (0%), while in the UC group, 88 out of 161 samples (54.7% showed positive expression. They found that expression of VEGF was associated with both locally confined and invasive tumors indicating that VEGF expression might be a predictive marker for the degree of invasiveness and thus can be of prognostic value. Bian et al.42 found a sensitivity of VEGF expression in urinary samples of 69% and a specificity of 88% for detection of UC. Eissa et al.91 conducted a study in which urinary samples were taken from 120 patients with bladder cancer, 54 patients with benign urological disorders and 55 healthy volunteers. The study revealed a sensitivity of 76.7% and a specificity of 61.5% for VEGF in the detection of UC. In conclusion, VEGF seems to play an important role in tumor progression and the identification of VEGF in urinary samples is mainly of prognostic value.

Epidermal growth factor receptor

Members of the epidermal growth factor receptor (EGFR) family, a type I tyrosine kinase growth factor receptor, are involved in various forms of cancers and serve as prognostic markers or therapeutic targets.92 Activation of EGFR leads to a wide variety of biological responses such as proliferation, differentiation, migration, modulation of apoptosis, invasion, and metastasis, leading to the progression of many tumors, including UC.93 Kassouf et al.94 found that EGFR expression predicts disease progression. In this study the authors did a UC tissue array from 248 archival paraffin blocks. High EFGR was associated with nonpapillary, high-grade and invasive tumors and may help select patients with UC for more aggressive therapy. A study with archival tissue by Litlekalsoy et al.95 showed a time dependent pattern of biological features in UC. In the 1930s these tumors tended to have a high proportion of high molecular weight cytokeratin and EGFR-positive cases, combined with more metastases and shorter life span. Seventy years later there is a tendency for the opposite pattern indicating the different environmental and carcinogenic influences that account for the development of bladder cancer.

In conclusion, the usage of growth factors bears promise, but it is mainly study-based and prognostic. No large trials have been conducted yet, therefore no clinically usable urinary markers have been identified.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

The development of urinary markers for the detection of bladder cancers is a dynamic field. New molecular markers are being developed and are under investigation. Our knowledge of complex cellular pathways is growing as the sequencing of the human genome molecular medicine evolves. An important discovery in biological research is the noncoding RNA (ncRNA). Novel ncRNA are providing new insights into gene regulation. Because the readout from next generation sequencers is quantitative, ncRNA characteristics will include detecting expression level changes that correlate with changes in environmental factors, with disease onset and progression.96 Technologies like proteomics and genomics will help us to find new and promising markers that in time will be ready to use in a clinical setting.

In current practice cytology has good sensitivity for detecting high-grade tumors and specificity is high. Detection of low-grade tumors is poor, but the clinical relevance of this is probably limited. Furthermore, cytology is highly operator dependent. So the perfect marker should have high sensitivity and high specificity, must have no interobserver variability and must be easy to perform. Overall, almost all mentioned urinary markers are better than cytology with regard to sensitivity but they do score lower in specificity. Finally, the perfect marker must have a good financial cost benefit ratio, which has not been studied sufficiently yet.

The perfect test is an ‘on bench’ test that gives result after a few minutes. The NMP22 BladderChek assay is such a test with somewhat better sensitivities than cytology, and is easy to perform. BTA-stat is also an ‘on bench’ test but scores less than cytology. Many other tests have a high workload and are more expensive. Telomerase, BTA-TRAK and CYFRA21-1 are influenced by benign urological conditions, which influence specificity and make these tests less usable in daily clinical practice. CK20 has been identified as a marker for progression but is best interpreted in combination with FGFR3, which is also used as a prognostic marker. The value of immunocyt test is also limited by its specificity and variability. MSA is promising because of its good sensitivity/specificity and its ability to also detect low-grade/stage tumors, but the test is complicated. Survivin and HA-HA-ase also have good sensitivity and specificity, but for both further studies are needed.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References

Currently, none of the previously discussed markers can guide us in diagnosis or surveillance, nor guide us in lowering the frequency of urethrocystoscopy. Molecular science is a dynamic field, however, and the future of marker development is bright as new techniques emerge; although the perfect marker is still to be found.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Molecular assays
  5. Molecular markers in urothelial cancers
  6. Other nuclear matrix proteins
  7. Cytokines
  8. Discussion
  9. Conclusion
  10. References