The standard of care for detection and surveillance of bladder cancer consists of cytology and cystoscopy, but both examinations suffer from many limitations, including issues related to accuracy, invasiveness, and cost. Several noninvasive methods for detection and surveillance of bladder cancer have been developed and urine-based biomarkers seem the most promising. This nonsystematic review critically analyzes the commercially available biomarkers and highlights some upcoming investigational biomarkers. To date, none of these biomarkers has sufficient validation to serve as a reliable alternative to cystoscopy.
Key points
- •
Due to its high rate of recurrence, bladder cancer (BC) requires a close follow-up that includes regular cytologic and cystoscopy examinations and leads to expensive lifetime health care expenditures.
- •
Urinary cytology has a low sensitivity especially for low grade tumors, whereas cystoscopy remains an invasive examination. Therefore, urine-based biomarkers should be considered good alternatives for the detection and follow-up of BC.
- •
Many biomarkers have shown a higher sensitivity than cytology. Most of them, however, failed to reach its specificity. A combination of biomarkers may increase their performance.
- •
A standardization of the techniques used for their detection followed by multicenter and prospective analysis needs to be performed before any assessment in large controlled clinical trials.
Introduction
With an estimated 74,000 new cases and 16,000 deaths for 2015, BC is the fifth most frequent malignancy in the United States. Urothelial carcinoma of the bladder (UCB) constitutes the most common histologic type and is the dominant histology in more than 90% of cases of BC. Approximately 75% of newly diagnosed BCs are non–muscle-invasive BCs (NMIBCs), of which 70% are Ta, 20% T1, and 10% carcinoma in situ. These lesions are usually treated with transurethral resection of the bladder with or without intravesical instillation therapies according to guidelines. The remaining 25% of BCs are muscle-invasive BCs (MIBCs) and the standard of care is radical cystectomy and bilateral lymph node dissection with or without perioperative chemotherapy. Although the prognosis of MIBC is poor (5-year mortality rate of 38%), the main issue with NMIBCs is that more than half of the patients experience disease recurrence within a period of 5 years and up to 30% of them experience disease progression to an MIBC despite therapy with curative intent.
This high recurrence rate is the primary reason that BCs have the highest lifetime treatment cost per person of all cancers. After initial treatment, patients with NMIBC are committed to a lifelong surveillance to identify recurrence early, with the goal of preventing disease progression to invasive disease. Depending on initial stage and grade, surveillance is of varying intensity but the current recommendation for high-grade tumors is cystoscopy and voided urine cytology every 3 months for 2 years, then every 6 months for 5 years, then yearly. Unfortunately, each recurrence restarts the scheduling scheme such that with 50% to 70% recurrence for high-grade tumors, many patients have ever more cystoscopic procedures.
With a sensitivity of 90%, standard white light cystoscopy is the gold standard for detection of BC, but it remains an invasive and costly examination, limiting the compliance of patients for the follow-up. Voided urine cytology is a highly specific test (99% specificity) but is limited by its low sensitivity (34%), especially in low-grade tumors and by interobserver variations.
Thus, to improve the management and the quality of life of patients with BC and to decrease the morbidity associated with current diagnostic and follow-up tests, many investigators have searched for a noninvasive, highly sensitive and specific marker of BC. Because urine is in contact with BC and can be collected noninvasively and in large amounts, urine-based assays are a natural and promising source for these biomarkers.
The assessment of a good biomarker should follow international guidelines. Guidelines edited by the International Bladder Cancer Network define biomarkers according to their clinical use: detection (screening and assessment of patients with hematuria) and follow-up of patients with BC. For each purpose, the required characteristics of biomarkers are different. To avoid unnecessary investigations and limit cystoscopies, a good diagnostic biomarker should have a low false-positive rate, whereas a good surveillance marker should have a high sensitivity and negative predictive value (NPV).
To date, 6 tests (BTA stat [Polymedco, Cortlandt Manor, New York], BTA TRAK [Polymedco], NMP22 BC test kit [Matritech, Newton, Massachusetts], NMP22 BladderChek Test [Alere, Waltham, MA], uCyt+ [Scimedx, Denville, New Jersey], and UroVysion Bladder Cancer Kit [Abbott Molecular, Des Plaines, Illinois]) have been approved by the Food and Drug Administration (FDA) and are commercially available for clinical use. A multitude of newer markers, however, using genetic testing are currently undergoing validation and have the potential to change clinical practice.
This nonsystematic review summarizes the current data on commercially available and emerging urinary biomarkers for detection and surveillance of BC. Screening for BC is not discussed because this is a complex subject requiring its own review.
Introduction
With an estimated 74,000 new cases and 16,000 deaths for 2015, BC is the fifth most frequent malignancy in the United States. Urothelial carcinoma of the bladder (UCB) constitutes the most common histologic type and is the dominant histology in more than 90% of cases of BC. Approximately 75% of newly diagnosed BCs are non–muscle-invasive BCs (NMIBCs), of which 70% are Ta, 20% T1, and 10% carcinoma in situ. These lesions are usually treated with transurethral resection of the bladder with or without intravesical instillation therapies according to guidelines. The remaining 25% of BCs are muscle-invasive BCs (MIBCs) and the standard of care is radical cystectomy and bilateral lymph node dissection with or without perioperative chemotherapy. Although the prognosis of MIBC is poor (5-year mortality rate of 38%), the main issue with NMIBCs is that more than half of the patients experience disease recurrence within a period of 5 years and up to 30% of them experience disease progression to an MIBC despite therapy with curative intent.
This high recurrence rate is the primary reason that BCs have the highest lifetime treatment cost per person of all cancers. After initial treatment, patients with NMIBC are committed to a lifelong surveillance to identify recurrence early, with the goal of preventing disease progression to invasive disease. Depending on initial stage and grade, surveillance is of varying intensity but the current recommendation for high-grade tumors is cystoscopy and voided urine cytology every 3 months for 2 years, then every 6 months for 5 years, then yearly. Unfortunately, each recurrence restarts the scheduling scheme such that with 50% to 70% recurrence for high-grade tumors, many patients have ever more cystoscopic procedures.
With a sensitivity of 90%, standard white light cystoscopy is the gold standard for detection of BC, but it remains an invasive and costly examination, limiting the compliance of patients for the follow-up. Voided urine cytology is a highly specific test (99% specificity) but is limited by its low sensitivity (34%), especially in low-grade tumors and by interobserver variations.
Thus, to improve the management and the quality of life of patients with BC and to decrease the morbidity associated with current diagnostic and follow-up tests, many investigators have searched for a noninvasive, highly sensitive and specific marker of BC. Because urine is in contact with BC and can be collected noninvasively and in large amounts, urine-based assays are a natural and promising source for these biomarkers.
The assessment of a good biomarker should follow international guidelines. Guidelines edited by the International Bladder Cancer Network define biomarkers according to their clinical use: detection (screening and assessment of patients with hematuria) and follow-up of patients with BC. For each purpose, the required characteristics of biomarkers are different. To avoid unnecessary investigations and limit cystoscopies, a good diagnostic biomarker should have a low false-positive rate, whereas a good surveillance marker should have a high sensitivity and negative predictive value (NPV).
To date, 6 tests (BTA stat [Polymedco, Cortlandt Manor, New York], BTA TRAK [Polymedco], NMP22 BC test kit [Matritech, Newton, Massachusetts], NMP22 BladderChek Test [Alere, Waltham, MA], uCyt+ [Scimedx, Denville, New Jersey], and UroVysion Bladder Cancer Kit [Abbott Molecular, Des Plaines, Illinois]) have been approved by the Food and Drug Administration (FDA) and are commercially available for clinical use. A multitude of newer markers, however, using genetic testing are currently undergoing validation and have the potential to change clinical practice.
This nonsystematic review summarizes the current data on commercially available and emerging urinary biomarkers for detection and surveillance of BC. Screening for BC is not discussed because this is a complex subject requiring its own review.
Material and method
The authors performed a Pubmed/Medline search on articles published in English from 2000 to June 2015 using a combination of the following keywords: bladder cancer, urothelial carcinoma, transitional cell carcinoma, urine, urinary biomarker, marker, surveillance, detection, diagnosis, follow-up, recurrence, and progression.
Commercially available biomarkers
Food and Drug Administration–Approved Biomarkers
Nuclear matrix protein 22
Nuclear matrix proteins (NMPs) are a family of proteins that play an important role in the structural framework of nucleus and are involved in every step of its function, from DNA replication to regulation of gene expression. NMP22 is specifically involved in mitosis by enabling a correct distribution of chromatin to daughter cells. The urinary concentration of NMP22 is 5-fold higher in patients with BC compared with healthy patients.
Two assays have been developed to detect NMP22 in voided urine. The original assay is the NMP22 BC test kit, a laboratory-based quantitative sandwich ELISA test using 2 antibodies. The NMP22 BladderChek Test is a qualitative immunochromatic assay designed as a point of care (POC) test. A few drops of urine on the cartridge containing NMP22 detection and reporter antibodies provide results within 30 minutes. Both tests have been approved by the FDA for BC surveillance, and the NMP22 BladderChek Test is also approved for detection of BC in patients at risk or presenting suspicious symptoms.
Pooled data analysis, including 41 studies and enrolling 13,885 patients, reported that NMP22 outperformed cytology with a sensitivity of 68% versus 44%. This high sensitivity was mainly due to a better detection rate of low-grade tumors compared with cytology. But NMP22 failed to reach the level of cytology for specificity (79% vs 95%) due a high rate of false-positive results. Because NMP22 is a ubiquitous nuclear protein, any aggression of the urinary epithelium (infection, inflammation, hematuria, urolithiasis, or instrumentation) can increase the release of NMP22 in the urine.
The clinical accuracy of NMP22 BladderChek has been evaluated in 2 large multicenter studies conducted by Grossman and colleagues. As a detection tool in patients with hematuria, NMP22 had a better sensitivity than urinary cytology (56% vs 16%) but its specificity remained lower (86% vs 99%). In a surveillance setting, the sensitivity and specificity of NMP22 were 50% and 87%, respectively. But in combination with cystoscopy, NMP22 increased significantly the detection rate of recurrence, up to 99% compared with cystoscopy alone (91%). The lower sensitivity of NMP22 test in a surveillance setting could be explained by the larger size and more advanced stage of tumors at the time of diagnosis compared with those detected during follow-up.
Tested in a large cohort of 1328 patients who were referred with hematuria to urologists, the overall positive predictive value (PPV) was 20% and the NPV was 97%.
The reliability and clinical utility of NMP22 have been questioned due to its low specificity compared with cytology and because the initial studies were performed with the laboratory-based assay that prevented the heterogeneity of widespread application. Decision curve analyses have suggested that its clinical benefit could be in decision making between immediate and delayed cystoscopy depending on a clinician’s threshold for conduction of cystoscopy. When compared with cytology-based nomograms, the accuracy of NMP22-based nomograms was higher (area under the curve–receiver operating characteristic curve [AUC-ROC] 82% versus 75%, P = .006).
In the detection setting, the use of a nomogram incorporating clinical factors (age, gender, and smoking) and BladderChek was validated prospectively in a multicenter cohort of 381 patients with a predictive accuracy of the BC detection nomogram of 80%.
At this time there is no current test to risk stratify patients with hematuria for referral to urology and there is a significant problem associated with lack of referral of patients with hematuria to urology. There is a potential role for a urine marker to benefit in this clinical scenario and validation studies, such as the one performed for BladderChek, are needed prior to incorporation into clinical care.
Bladder tumor antigen
The bladder tumor antigen (BTA) is a human complement factor H-related protein produced by BC cells and is similar to human complement factor H. By interrupting the complement cascade activation, BTA confers a selective growth advantage and allows tumor cells to evade the host immune system.
The BTA test exists in 2 assays, both designed to detect BTA in voided urine. BTA TRAK is a quantitative, laboratory-based ELISA assay, whereas BTA stat is a qualitative and immunochromatic POC device. Its design enables its use in a clinical setting. Both tests have been approved by FDA only for surveillance in complement of cystoscopy.
The reported overall sensitivity and specificity of BTA TRAK were 66% and 65%, respectively, whereas BTA stat had a sensitivity of 70% and a specificity of 75%. For both tests, sensitivity improved with increasing histologic stage and grade, but specificity remained lower than cytology and the improvement of accuracy in sensitivity for low-grade, low-stage tumors remained modest.
These high rates of specificity have to be balanced by the fact that many studies excluded patients with benign genitourinary conditions. When including these patients, specificity dropped to 56%. Most of the false-positive results were due to hematuria, benign prostatic hyperplasia, urolithiasis, infection, inflammation, history of bacille Calmette-Guérin (BCG) instillations, and bowel interpositions. BTA has today a limited clinical value with only few centers using it for any clinical decision making.
ImmunoCyt/uCyt+
The uCyt+ assay (formerly called ImmunoCyt) is a combination of cytology and immunofluorescence. It detects exfoliated BC cells in the urine by using 3 fluorescent monoclonal antibodies targeting 3 specific antigens of BC cells: M344 is a high-molecular-weight form of carcinoembryonic antigen, and LDQ10 and 19A11 are bladder tumor cell–associated mucins. The test requires trained cytopathologists and is performed under microscopy. A large number of exfoliated cells (more than 500 per slide) is required to perform an accurate test. The test is scored positive when the presence of either 1 red or green fluorescent cell is observed, but the manufacturer recommends that all positive cells should be correlated to morphology. The test is approved by the FDA for monitoring of patients with a history of BC, as an adjunct to cystoscopy.
In a systematic review, including 10 studies and 4199 patients, the reported overall sensitivity was 84% and specificity was 75%. In low-grade and low-stage tumors, uCyt+ had a superior sensitivity compared with cytology alone. In a study including 2217 patients, when combined with cytology, uCyt+ reached an overall sensitivity of 73%, with 59% for grade 1, 77% for grade 2, and 90% for grade 3 tumors, but specificity of combined assays remained lower than cytology alone (72% vs 98%, respectively). In a study including 870 patients, the NPVs of cytology, uCyt+, and both analyses were 88%, 93%, and 95%, respectively, and the PPVs were 70%, 26%, and 29%, respectively.
As a cell-based assay, uCyt was less impacted by hematuria and inflammatory conditions compared with other urinary assays. In a multicenter prospective study that enrolled 1182 patients without a history of BC and presenting with painless hematuria, uCyt+ was a strong predicator of BC. Decision curve analyses on multivariable models, including uCyt+, achieved the highest predictive accuracy (91%). Nevertheless, the limited evidence and the user dependency of this assay had led to infrequent use in clinical care.
UroVysion
The UroVysion Bladder Cancer Kit is a multitarget fluorescence in situ hybridization (FISH) assay that identifies the most common urothelial carcinoma–related chromosomal alterations in exfoliated cells in urine: aneuploidy for chromosomes 3, 7, and 17 and the loss of the 9p21 locus of the p16 tumor suppressor gene. With a fluorescent microscope, the cytopathologist needs to count the 4-color fluorescent signals that assess the copy number of each target in the nuclei. To date, there are no uniform criteria for positive UroVysion, but the test is generally considered positive when a minimum of 25 abnormal cells is observed, including at 4 cells with a gain of 1 or more chromosome, or 12 or more cells with an homozygous loss of the 9p21 locus. The test is FDA approved for use in conjunction with current standard procedures for detection of BC in patients with hematuria and surveillance of patients with history of BC.
A pooled data analysis performed on 14 studies involving 2477 FISH tests found that UroVysion outperformed cytology (AUC-ROC 87% vs 63%) and had overall sensitivity of 72%, but a lower specificity than cytology (83%). When excluding Ta tumors, the sensitivity of UroVysion reached 86% compared with 61% for cytology, but the overall test performance almost disappeared when excluding this population, suggesting that UroVysion has a better sensitivity in low-grade tumors.
Due to its laboratory cell-based nature, UroVysion accuracy is dependent on several technical aspects, such as laboratory staff experience with performing FISH and sample quality with a sufficient number of tumor cells. The use of automatic scanning systems combining FISH and cellular morphology analysis improves the accuracy for BC detection.
Several follow-up studies reported that almost half of the patients with initial false-positive FISH tests and negative cystoscopy results experienced disease recurrence within the year after the test, suggesting that the detection of chromosomal abnormalities anticipated the diagnostic of recurrence by cystoscopy or urinary cytology. Therefore, an abnormal UroVysion test could constitute an accurate surveillance assay by anticipating disease recurrence. Some other studies promoted a reflex FISH in cases of atypical cystoscopy or cytology to assess tumor recurrence. This type of strategy may reduce number of unnecessary biopsies and may be cost effective. Finally, UroVysion could also be useful in monitoring patients treated with intravesical BCG. Further studies are necessary to validate these indications. The cost-effectiveness of such labor-intensive tests need to be assessed in different health care settings.
Non–Food and Drug Administration–Approved Biomarkers
CxBladder
Cxbladder Detect (Pacific Edge, Hummelstown, PA) is based on the detection 4 mRNAs significantly increased in voided urine and patients with BC (IGFBP5, HOXA13, MDK, and CDK1) and another mRNA (CXCR2) that is associated with nonmalignant inflammatory conditions to reduce false-positive results. The expression of these mRNAs is assessed with real-time reverse transcription polymerase chain reaction (RT-qPCR). In a prospective study, including 485 patients with hematuria and without history of BC, CxBladder reached a sensitivity of 82% when the cutoff was prespecified to give a specificity of 85%. CxBladder seemed to be able to distinguish between low-grade Ta tumors and other detected urothelial carcinoma with a sensitivity of 91% and a specificity of 90%. These data need to be confirmed by further studies.
Survivin
Survivin is an antiapoptotic protein that is almost exclusively expressed by malignant epithelium. Several techniques and assays have been used to detect mRNA or the protein level of survivin, but the commercially available assay is a dot-blot technique (BioDot assay, Fujirebio Diagnostics [Malvern, PA]).
A meta-analysis, including 2051 subjects, reported a sensitivity of 77% and a specificity of 92% and an AUC-ROC of 94%. Nevertheless, these results need to be balanced by the lack of standardization of assay and cutoff values that yield the interpretation of the data difficult.
Cytokeratin fragment 21.1
Cytokeratin is a family of marker of epithelium differentiation and some members have been related to BC. Cytokeratin fragment 21.1 (CYFRA 21.1) is an ELISA assay that detects fragments of cytokeratin 19.
In a pooled data analysis, including 2495 patients, the sensitivity was 82% and the specificity was 80%. The AUC-ROC was 87%. When including patients with benign conditions, such as urolithiasis, infection, history of BCG, and radiotherapy, the high rate of false-positive results led to a lower specificity, making CYFRA 21.1 not useful as a surveillance tool anymore.
Bladder cancer rapid test
Cytokeratin 8 and 18 can be detected by the urinary BC test (UBC, IDL Biotech, Bromma, Sweden) with a POC assay (UBC Rapid test [IDL Biotech]) or an ELISA assay. The POC assay can provide qualitative results within 10 minutes, whereas ELISA provides quantitative results.
In a study including 112 patients, the reported sensitivity and specificity for the UBC Rapid test were both 64%. When including potential false-positive confounders in cohorts, such as other urinary tract malignancies or benign conditions, the sensitivity was 79% and the specificity 49%. In these cases, the UBC Rapid test performed worse than BTA tests. The false-positive rate reached 20% for patients with benign urologic diseases and 44% for patients with other urinary tract malignancies. Most of the side-by-side comparisons with other biomarkers were not in favor of the UBC Rapid test but, when combined with the POC reader instead of visual reading, the diagnostic accuracy seemed to improve.
Table 1 summarizes commercially available biomarkers.
| Marker | Assay Type | Sensitivity (%) | Specificity (%) | Food and Drug Administration Approved | Ref. |
|---|---|---|---|---|---|
| Cytology | Giemsa or hematoxylin-eosin staining | 34 | 99 | Diagnostic and follow-up | |
| NMP22 | ELISA | 40 | 99 | Follow-up | |
| NMP22 | POC device | 68 (62–74) | 79 (74–84) | Diagnosis and follow-up of high risks | |
| BTA stat | Dipstick immunoassay | 70 | 75 | Diagnosis and follow-up | |
| BTA TRAK | Sandwich ELISA | 66 | 65 | Diagnosis and follow-up | |
| uCyt+ | Immunocytochemistry | 84 (77–91) | 75 (68–83) | Follow-up in adjunct to cystoscopy | |
| UroVysion | Multicolored and multiprobed FISH | 72 (69–75) | 83 (82–85) | Diagnosis and follow-up | |
| CxBladder | RT-qPCR | 82 | 85 | Not approved | |
| Survivin | BioDot test | 77 (75–80) | 92 (90–93) | Not approved | |
| CYFRA 21.1 | Immunoradiometric assay or ELISA | 82 (70–90) | 80 (73–86) | Not approved | |
| UBC test | Sandwich ELISA or POC | 64 | 64 | Not approved |
Investigational biomarkers
Protein-Based and Cell-Based Biomarkers
Apoptosis markers
Soluble Fas (sFas) isoforms are antiapoptotic proteins produced and released by BC cells, protecting them from host antitumor immunity and are measurable by ELISA. Urinary levels of sFas have been shown an independent predictor of BC recurrence. In a study, including 191 patients, reported sensitivity and specificity for sFas were 88% and 89%, respectively. When compared with NMP22, sFas seemed a better predictor of BC and invasiveness with an AUC-ROC of 76%, outperforming NMP22 (70%).
Clusterin is a multifunctional secretory glycoprotein that has a potential role in development and progression of several human cancers and is measurable by ELISA. Urinary levels of clusterin were significantly higher in patients with BC. Reported sensitivity rates were between 70% and 87%. Reported specificity was between 83% and 97%.
Angiogenesis markers
Vascular endothelial growth factor (VEGF) is a key mediator of angiogenesis produced by BC cells and measurable by ELISA in voided urine. Mean levels of VEGF were significantly associated with presence of BC and increased with tumor stage. The reported sensitivity and specificity ranged from 68% to 83% and from 62% to 93%, respectively. The AUC-ROC was 89%.
Interleukins (ILs) are small signaling proteins secreted by white blood cells and involved in the inflammatory process of the immune system. They are measurable by ELISA. Urinary levels of IL-8 were significantly higher in patients with BC and were correlated with tumor stage. In a detection setting, IL-8 showed a high accuracy with an AUC-ROC at 79%. In a post-BCG surveillance setting, urinary levels of IL-8 were greater in patients who experienced disease recurrence : at a cutoff of 112 pg/mL, IL-8 measured 2 hours after BCG instillation predicted recurrence with a sensitivity of 53%, specificity of 89%, PPV of 73%, and NPV of 77%. IL-6 was also an independent predictor of BCG response, and the ratio IL-6/IL-10 has been shown to predict both recurrence after BCG response and recurrence in patients at intermediate risk, with a sensitivity of 83% and a specificity of 76%.
Although angiogenesis and inflammation are definite hallmarks of cancer, they are relatively nonspecific and, therefore, unlikely to help in BC diagnosis or surveillance.
Proliferation and invasion
Telomerase is a ribonucleoprotein enzyme that synthetizes telomeres (repeated sequences of TTAGGG) at the ends of chromosomes to ensure genome stability. Several malignant cell types, including BC cells, acquire immortality by hyperactivating telomerase. Telomerase activity can be measured by different assays: the telomeric repeat amplification test is a PCR-based technique, or measuring the activity by RT-qPCR of human telomerase RNA the telomerase reverse transcriptase (hTERT). The most accurate method seemed to be hTERT, with an overall reported sensitivity from 75% to 96%, specificity from 69% to 96%, NPV of 91%, and PPV of 96%. From grades 1 to 3, reported sensitivity was 52%, 80%, and 94%, respectively. The assays still need to be standardized and validated before widespread use. Moreover, the lack of specificity of telomerase activity makes it a biomarker of little value with a high rate of false-positive.
Hyaluronic acid (HA) is a glycosaminoglycan constitutive of the extracellular matrix and is involved in cell adhesion and proliferation. It is degraded by hyaluronidase (HAase) into small fragments that promote angiogenesis. Both components are increased in urine of patients with BC. HA-HAase is measurable by ELISA, and HAase can be assessed by RT-qPCR. When combined, the performance of the test was better than either test alone, reaching a sensitivity of 92%, specificity of 85%, accuracy of 88%, PPV of 64% to 92%, and NPV of 67% to 91%. The levels of HA/Hases were correlated with tumor grade. Further research is needed to assess and establish the clinical relevance of this biomarker.
Fibronectin is a structural glycoprotein widely present in cells, plasma, and extracellular tissue matrix that is implicated in cell migration and adhesion. When tumors are present, the components of the extracellular matrix are degraded by proteases resulting from metastatic or invasive conditions. In a meta-analysis, including 5 studies involving 649 patients and 291 controls, urine levels of fibronectin had a pooled sensitivity of 81% and specificity of 80%. The AUC-ROC was 86%. These results are interesting but early because the clinical scenario needs to be identified and tested.
CD44 antigen is a cell-surface glycoprotein involved in cell adhesion, proliferation, and migration. Among CD44 variants, the variant 6 (CD44v6) expression is measurable in the urine by RT-qPCR and was significantly increased in patients with BC, with a reported sensitivity between 50% and 86%, specificity between 72% and 79%, and PPV of 78%. Here also, a valid, reliable, and reproducible assay is far from development.
Metabolomics
Like genomics and proteomics, metabolomics is another “omics” science that has emerged in recent years. It aims to identify a biological signature of BC based on the analysis of metabolites produced in an abnormal quantity by BC cells compared with normal cells. Various analytical platforms based on liquid chromatography and mass spectrophotometry are currently used to analyze these metabolites. More than 10,000 compounds have been identified so far, but when reduced to a panel of 3 to 15 metabolites, the reported sensitivity, specificity, and AUC-ROC are 91% to 100%, 93% to 100%, and 90% to 94%, respectively. Their detection seems correlated with cancer-specific survival. The domain of research is preliminary and further studies are needed to assess the clinical relevance of such techniques.
Table 2 summarizes investigational biomarkers.
| Marker | Assay Type | Sensitivity (%) | Specificity (%) | Ref. |
|---|---|---|---|---|
| sFas | ELISA | 88 | 89 | |
| Clusterin | ELISA | 70–87 | 83–97 | |
| VEGF | ELISA | 68–83 | 62–93 | |
| Telomerase | RT-qPCR PCR | 75–96 | 69–96 | |
| HA HAase | ELISA RT-qPCR | 92 | 85 | |
| Fibronectin | IEMA | 81 | 80 | |
| CD44v6 | RT-qPCR | 50–86 | 72–79 | |
| AURKA | RT-qPCR | 84 | 65 | |
| FGFR3 | Snapshot analysis | 25–58 | 99 |
Gene-Based Biomarkers
Aurora A kinase
Aurora A kinase (AURKA) is a serine/threonine kinase implicated in the regulation of gene stability during the mitosis. Overexpression of AURKA gene can be assessed by FISH, with a sensitivity of 87%, specificity of 97%, and AUC-ROC of 94%. It can also be measured through AURKA mRNA expression with RT-qPCR, providing an overall sensitivity of 84% and specificity of 65%. When compared with cytology, accuracy of AURKA was particularly evident in patients with low-grade tumors, with a predictive accuracy of 73 versus 59%. This biomarker holds much promise, and further studies are awaited.
Fibroblast growth factor 3 receptor
Mutations in fibroblast growth factor 3 receptor (FGFR3) are present in more than 50% of voided urine of BC patients and are more common in low-grade (70%) and low-stage BC (60%). A multiplex PCR (SNaPshot [Applied Biosystems, Foster City, CA]) has been used to assess it. FGFR3 mutations were an independent factor of BC recurrence. In a surveillance setting for low-grade tumors, the sensitivity of the assay was 58%, which was higher than cytology. Moreover, when combined with cytology, the sensitivity reached 76%. In a detection setting in patients with hematuria, sensitivity, specificity, PPV, and NPV were 25%, 99%, 17% and 99%, respectively. A cost-effectiveness study has shown that surveillance in which cystoscopy was partly replaced by mutation analysis of urine seemed safe, effective, and cost effective. This biomarker has high potential because it may have several roles, including cancer detection, therapeutic target, and tool for monitoring for treatment response.
Microsatellite/loss of heterozygosity detection
Microsatellites are highly polymorphic short tandem DNA repeats found through the genome resulting from a failure of the DNA mismatch repair and playing an important role in tumor cell transformation. This loss of heterozygosity could be a marker of carcinogenesis. It has been established that microsatellite changes in urine samples matched DNA extracts from tumor tissue. The reported overall sensitivity ranged from 79% to 84%, increasing with tumor grades 1 to 3 from 75% to 96%; specificity ranged from 85% to 100%. The accuracy of the test is too low to use it in a surveillance setting (sensitivity 58% and specificity 73%) and would not be cost effective. This biomarker may have a specific role in upper tract urothelial carcinoma rather than BC.
DNA methylation
Table 3 summarizes some of the genes investigated for methylation epigenetic changes. In a detection setting, reported sensitivity ranged from 65% to 100% and specificity from 77% to 100% (see Table 3 ). Few studies explored DNA methylation in a surveillance setting. In a study including 184 patients, Reinert and colleagues reported a sensitivity of 82% to 89% and a specificity of 94% to 100% when studying a panel of 6 genes. Further studies, with standardized techniques should be performed to assess the role of DNA methylation detection in voided urine and its potential integration in detection and follow-up of patients with BC. Methylation biomarkers could also serve as a target for epigenetic modifiers that would enhance the efficacy of other therapies.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
