Lozano et al. BMC Urology (2020) 20:99 https://doi.org/10.1186/s12894-020-00670-x RESEARCH ARTICLE Open Access Current status of genetic urinary biomarkers for surveillance of non-muscle invasive bladder cancer: a systematic review F. Lozano1,2* , C. X. Raventos1, A. Carrion1, E. Trilla1 and J. Morote1,2 Abstract Background: Genetic biomarkers are a promising and growing field in the management of bladder cancer in all stages. The aim of this paper is to understand the role of genetic urinary biomarkers in the follow up of patients with non muscle invasive bladder cancer where there is increasing evidence that they can play a role in avoiding invasive techniques. Methods: Following PRISMA criteria, we have performed a systematic review. The search yielded 164 unique articles, of which 21 articles were included involving a total of 7261 patients. Sixteen of the articles were DNA based biomarkers, analyzing different methylations, microsatellite aberrations and gene mutations. Five articles studied the role of RNA based biomarkers, based on measuring levels of different combinations of mRNA. QUADAS2 critical evaluation of each paper has been reported. Results: There are not randomized control trials comparing any biomarker with the gold standard follow-up, and the level of evidence is 2B in almost all the studies. Negative predictive value varies between 55 and 98.5%, being superior in RNA based biomarkers. Conclusions: Although cystoscopy and cytology are the gold standard for non muscle invasive bladder cancer surveillance, genetic urinary biomarkers are a promising tool to avoid invasive explorations to the patients with a safe profile of similar sensitivity and negative predictive value. The accuracy that genetic biomarkers can offer should be taken into account to modify the paradigm of surveillance in non muscle invasive bladder cancer patients, especially in high-risk ones where many invasive explorations are recommended and biomarkers experiment better results. Keywords: Biomarkers, Surveillance, Bladder, Genetic * Correspondence: flozano@vhebron.net 1 Urology Department, Vall d’Hebron University Hospital, Pg. Vall d’Hebron 119-129, 08035 Barcelona, Spain 2 Universitat Autònoma de Barcelona, Barcelona, Spain © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Lozano et al. BMC Urology (2020) 20:99 Page 2 of 11 Background Methods Bladder cancer (BC) is the fifth most common neo- FLP performed a bibliographic search of Medline plasm worldwide, with more than 54,000 new cases (http://www.ncbi.nlm.nih.gov), Embase (http://www. estimated per year in the United States alone [1] . BC embase.com) and the Cochrane library (http://www. is a heterogeneous tumour that is associated with cochrane.org) up to March 2020. MeSH terms used were very high economic costs and a substantial impact on Bladder cancer AND surveillance AND biomarkers patients’ quality of life owing to its characteristically AND DNA OR RNA OR methylation, yielding 2241 arti- high risk of recurrence and the complexity of follow cles. After that, two authors (FLP and CXR) screened all up [2]. Guidelines from the European Association of published original articles appearing in the above search Urology (EAU) and the American Urological Associ- for eligibility. Studies using genetic urinary biomarkers ation (AUA) suggest a combination of cystoscopy, for surveillance in non muscle invasive bladder cancer in cytology and imaging for the surveillance of patients humans were selected. Studies were excluded if they with non-muscle invasive bladder cancer (NMIBC) [3, were not original research papers, used a language other 4]. Cystoscopy is an invasive procedure that carries than English, had less than 20 patients or did not report the risks of painful micturition, urinary frequency and biomarker performance in terms of sensitivity, specifi- macroscopic haematuria of 50, 37 and 19%, respect- city, or area under the curve (AUC); or reported the per- ively [5], while cytology has a very low sensitivity, es- formance of genetic markers only in combination with pecially for low-grade tumours [6, 7] . other factors (clinical data or non genetic biomarkers), For this reason there has been an increase in re- yielding 164 articles. search over the past years into urinary biomarkers for After applying the PRISMA (Preferred Reporting Items the three scenarios of haematuria, diagnosis and sur- for Systematic Reviews and Meta-analyses) criteria nar- veillance. The role of these new tests is to increase rowed this down to 21 original articles (Fig. 1). the sensitivity and the specificity of the available After reviewing each of the selected articles using the gold-standard techniques, while sparing the patient REMARK checklist [10], we designed a QUADAS-2 the discomfort of an invasive test and its potential table to evaluate each study’s risk of bias and quality complications. Although many types of urinary bio- (Fig. 2). The main bias observed was reference to the markers have been investigated, biomarkers that use index test. Although we identified a 50% risk of bias, genetic materials such as DNA and RNA seem to be most of the articles reviewed met most of the QUADAS- the most promising due to their potential to identify 2 criteria [11], using the four considerations (patient se- a genetic signature. Such a signature would not only lection, index, reference and flux and timming) which prove useful in disease detection and follow-up but suggested that the studies were of moderately high also in the facilitation of more precise treatment by quality. avoiding unhelpful therapies that may delay the best We have also used the SWIM (Synthesis without oncological pathway. meta-analysis) reporting guideline [12] to clarify the The field of urinary biomarker research in BC is fo- reporting methodology of the article (Fig. 3). The cused on balancing a non-invasive, safe method with a evidence was reported in terms of sensitivity and cost-effective strategy that can be used to improve the specificity for each biomarker (for low grade and sensitivity of bladder tumour detection in the initial high grade tumors in case of studies that indicated phase of the disease and during patient follow-up, com- that subestratification). Area under de curve has pared with the current gold standard. been informed in eleven of the seventeen studies. Selecting a biomarker must be based on the given We have also reported negative and positive predict- scenario and follow the principles of the inter- ive value of the markers. Recurrence rate has been national guidelines [8, 9] . The current literature calculated using the positive cases (positive path- clearly differentiates between different biomarker ology) and the total number of samples. We have tests and characteristics depending on whether the prioritized articles identified as low risk of bias BC is low or high risk. For low-risk tumours, based on QUADAS-2 table to draw the conclusions marker-guided testing of lesions is suggested to de- of this review. We have also performed exploratory tect possible progression to high-risk tumours. For analyses to determine whether different study char- high-risk tumours, however, where early detection is acteristics varied the effects of the interventions. Al- the main objective, selection of high-sensitivity bio- most all the studies accepted as confirmed positive markers is recommended [9]. case if there is a pathology report. Some studies gen- The aim of this study is to analyse the current litera- erate artificial cohort. We examined whether this dif- ture for the use of genetic urinary biomarkers in the sur- ferent type of targeted behaviour modified, on veillance of NMIBC. average, the effect of the interventions. Lozano et al. BMC Urology (2020) 20:99 Page 3 of 11 Fig. 1 Flow chart following PRISMA criteria Results to yield better results for detecting recurrences (AUC In order to obtain a practical and visual description of 0.81 vs 0.44). When a Bayesian network analysis was the different studies, we divided the biomarkers by performed that combined variables and biomarkers, the method into DNA-based (Table 1) and RNA-based panel of markers generated a sensitivity of 85% and a (Table 2) tests. specificity of 86%. Van der Aa et al. [14] designed a mul- ticentre study to evaluate the clinical utility of MA in DNA tests (Table 1) low-grade tumours in combination with FGFR3 muta- DNA tests used for surveillance are based on microsatel- tions described previously [35]. The sensitivity in this lite analysis (MA). They are employed to detect loss of study was 58% and the specificity 73%, with a negative heterozygosity, gene methylation levels and gene muta- predictive value of 94%. tions in cells collected from urine. DNA methylation has been recognized to be important Microsatellite markers are highly polymorphic tandem in the developmental biology and cancer aetiology of repeat DNA sequences distributed throughout the gen- many neoplasms [36–38]. DNA methylation is an epi- ome and easily amplifiable by standard polymerase chain genetic marker that mainly affects CpG dinucleotides. reaction [34] . Rouprêt et al. [13] compared this bio- These dinucleotides are distributed throughout the gen- marker with methylation biomarkers in a comparative ome and usually have a normal methylation status. cohort study of 40 patients. In this study, MA appeared Hypermethylation of CpG dinucleotides in the promoter Lozano et al. BMC Urology (2020) 20:99 Page 4 of 11 Fig. 2 QUADAS 2 table evaluating quality of the studies Lozano et al. BMC Urology (2020) 20:99 Page 5 of 11 Fig. 3 Synthesis Without Meta-analysis (SWiM) reporting items Lozano et al. BMC Urology (2020) 20:99 Page 6 of 11 Table 1 DNA based biomarkers used in follow-up for non muscle invasive bladder cancer patients Reference Patients/ Recurrence rate Sensitivity(%) Specificity(%) NPV PPV AUC Method Markers samples (%) (%) Roupret et al 2008 40/40 38% 80 68 (microsat) 0.81 DNA Microsatellitea vs [13] (microsatellite) 8 (methyl) (microsat) PCR methylationb 86 86 0.44 (methylation) (combination) (methyl) 85 (combination) Van der Aa et al 228/815 10.3% 58 73 94 61- NA DNA Microsatellite + 2009 [14] 77 PCR FGFR3 mutation Zuiverloon et al 134/463 9.7% 58 NA 89 25 NA DNA FGFR3 mutationsc 2010 [15] PCR Reinert et al 2012 158/206 67.4% 87-94 28-47 55-78 72- 0.68-0.78 DNA Methylation d [16] 78 PCR Zuiverloon et al NA/94 69.1% 72.3 55.2 NA NA NA DNA Methylation genes 2012 [17] PCR APC_a , TERT_a , TER _b ,EDNRB Allory et al 2013 194/395 44.8% 19(FGFR3) 73 (TERT) NA NA NA DNA Gene mutations [18] 42(TERT) 90(FGFR3) PCR (TERT and FGFR3) 50(FGFR3+TERT) 71(FGFR3+TERT) Abern et al 2014 111/111 21.6% 75-79 63-71 92 37- 0.74 DNA Methylation genes [19] 42 (TWIST1) PCR TWIST1, NID2 0.68 (NID2) Su et al 2014 [20] 90/368 37.7% 80 97 NA NA 0.95 DNA Hyper and hypomethylated PCR genes (SOX1, IRAK3, L1-MET) Fantony et al 2015 126/126 25% 58-67 61-69 83-85 36- 0.66 DNA Methylation genes TWIST1, [21] 38 (TWIST1) PCR NID2 0.63 (NID2) Beukers et al 2016 NA/2191 64% 57 (LG) 59% LG NA NA NA DNA FGFR3 mutation, TERT [22] 72 (HG) PCR mutation and OTX1 methylation Roperch et al 2016 158/613 45.5% 94.5 75.9 98.5 NA 0.82 DNA FGFR3 mutation [23] 96 (HG) PCR +DNA methylation HS3ST2, SLIT2 and SEPTIN9 Van der Heijden NA/458 37.7% 90 31 82 50 0.74 DNA DNA gene Methylation et al 2018 [24] PCR (CFTR, SALL3, TWIST1) Witjes et al 2018 353/353 13% 68.2 88 95.1 44.8 0.82 DNA 15 DNA methylation genes [25] 92.6 (HG) 99.3( PCR (Epicheck®) HG) Springer et al 2018 322/322 58% 68 80 NA NA NA DNA 10 gen mutationse plus [26] 71 (HG) PCR detection of aneuploidy (UroSEEK®) D’Andrea et al 357/357 13.7% 67.3 88 94 47 85.9 DNA 15 DNA methylation 2019 [27] 88.9 (HG) 88(HG) 99 30 PCR genes (Epicheck®) (HG) (HG) Batista et al 2019 122/122 28% 73.5 73.2 NA NA NA DNA TERT promoter and [28] PCR FGFR3 mutations (Uromonitor®) LG low grade, HG high grade, NA not allowed a FGA (4q28), D4S171(4q35)), 5 (ACTBP2(5q14)), 9 (D9S162 (9p), IFNA (9p21)), 14 (MJD52(14q32)), 16 (D16S310 (16q21)) and 18 (D18S51 (18q21), MBP (18qter). b (RASSF1a (3p21.3),E-cadherin (16q22.1), APC (5q21), DAPK (9q22.1), MGMT (10q26), BCL2 (18q21.33), h-TERT (5p15.33), EDNRB (13q22), WIF-1 (12q14.3), TNFRSF25 (1p36.31), IGFBP3 (7p13)) c R248C and S249C (exon 7); G372C,S373C, Y375C, G382R, and A393E (exon 10); and K652M, K652T, K652E, and K652Q (exon 15) d EOMES, HOXA9, POU4F2, TWIST1, VIM, ZNF154 e FGFR3, TP53, CDKN2A, ERBB2, HRAS, KRAS, PIK3CA, MET, VHL, MLL and TERT promoter. Lozano et al. BMC Urology (2020) 20:99 Page 7 of 11 Table 2 RNA based biomarkers used in follow-up for non muscle invasive bladder cancer patients Reference Patients/samples Recurrence rate Sensitivity Specificity NPV PPV AUC Method Markers Sapre et al 2016 [29] 131/131 NA 88 48 75 63 0.74 miRNA PCR 6 miRNA signaturea Kavalieris et al 2017 [30] 736/1036 15.1% 92 NA 96 NA 0.73 mRNA PCR 5 genes mRNA expression (Cx Bladder Monitor® )b Lotan et al 2017 [31] 748/1016 14.8% 91 NA 96 NA NA mRNA PCR 5 genes mRNA expression 95 (HG) (Cx Bladder Monitor® )b vs NMP22 ELISA vs NMP22 BladderChek Pilcher et al 2018 [32] 140/155 30.7% 84 91 93 72 0.87 mRNA RT-PCR ABL1, CRH, IGF2, UPK1B, 100 (HG) ANXA10 (Xpert Bladder Cancer Monitor®) Wallace et al 2018 [33] 370/370 13.2% 73 77 92 44 0.87 mRNA RT-qPCR ABL1, CRH, IGF2, ANXA10, 83 (HG) UPK1B (Xpert Bladder Cancer Monitor®) HG high grade, NA not allowed a miR16, miR200c, miR205, miR21, miR221 and miR34a b IGFBP5, HOXA13, MDK, CDK1, CXCR2 regions of tumour suppressor genes can repress their cytology). Witjes et al. [25] evaluated a combination of transcription in human cells [39, 40]. Methylation status 15 methylated genes (Epicheck®), obtaining a sensitivity is one of the most studied biomarkers in the follow-up of 68.2% (92.6% for high-grade tumours) and a specifi- scenario because it is both chemically stable and quanti- city of 88%. D’Andrea et al. [27] published another mul- fiable [41]. Zuiverloon et al. [17] developed a retrospect- ticentric and independent study using the same test, ive four-step test, selecting methylation of the APC_a, supporting the sensitivity (67.3, 88.9% for high grade) TERT_a, TER_b and EDNRB genes as the combination and specificity (88%) described in the previous publica- providing a higher sensitivity and specificity (63.3 and tion by Witjes. Abern [19] studied the role of two meth- 58.3%, respectively) than other combinations investi- ylated genes, TWIST1 and NID2 based on Renard work gated in this study. Based on their previous study [42], [45] due to their high sensitivity and specificity for Reinert et al. evaluated the methylation of EOMES, urothelial carcinoma. They observed that TWIST1 HOXA9, POU4F2, TWIST1, VIM and ZNF154. Their methylation had better AUA than NID2 or the combin- study consisted of a first step, validating the markers and ation of both genes. They also showed that adjusting the establishing the cut-off levels, and a second step in the thresholds, the test had a sensitivity and specificity of 75 surveillance scenario excluding those patients who and 71%, respectively. Fantony et al. [21] published a showed no aberrant methylation of their tumour marker more recent multi-institutional study using the same genes. The authors reported a sensitivity of between 87 methylated genes, obtaining similar conclusions and re- and 94% and a specificity ranging from 43 to 67%. Com- sults of sensitivity (58–67%) and specificity (61–69%). In bining the different biomarkers did not improve the ac- this paper, prior BCG treatment for NMIBC reduced the curacy of the test [16]. Su et al. [20] tested six DNA accuracy of the test. methylation markers before building a model with Many of the gene mutations investigated are related to SOX1, IRAK3 and L1-MET as the best combination to the carcinogenesis of urothelial carcinomas, which are detect recurrences. Using this model they obtaining a among the most heterogeneous tumours [46]. One of sensitivity of 80% and a very high specificity of 97%. the most studied among these genes is fibroblast growth Roperch et al. [23] combined four different FGFR3 mu- factor receptor 3 (FGFR3), mutations of which are found tations and eighteen methylation markers based on the in almost 80% of the low-grade tumours and associated literature [43, 44]. Finally, they selected three of these with a good prognosis [35, 47]. markers (the genes HS3ST2, SLIT2 and SEPTIN9) for Zuiverloon et al. [15] evaluated this marker in non- combination with the FGFR3 mutations in a logistic re- high grade tumours, achieving a sensitivity of 58%. Beu- gression model, obtaining a sensitivity of 94.5% (96% in kers et al. [22] combined FGFR3 mutation with TERT high-grade tumours) and a specificity of 75.9%. Van der mutation and OTX1 gene methylation in a large pro- Heijden et al. [24] evaluated seven selected genes that spective European cohort study, obtaining a sensitivity are found at significantly increased levels in the urine of 57% for low-grade and 72% for high-grade BC. Allory sediment from patients with BC. After testing a training investigated the role of telomerase reverse transcriptase set, they selected the CFTR, SALL3 and TWIST1 genes (TERT) promoter mutations, frequently founded in for validation in a large series (458 samples) and ob- many other non urothelial tumors [48] in combination tained a sensitivity of 90% (96% in combination with with FGFR3 mutation [18]. This study showed that Lozano et al. BMC Urology (2020) 20:99 Page 8 of 11 combination of TERT and FGFR3 has higher sensitivity confirmed that cytology did not enhance diagnostic (50%) than TERT or FGFR3 individually. Moreover, accuracy. FGFR3 had higher specificity than TERT mutation. In a more recent multicentric study, Batista et al. Discussion [48] have developed a biomarker based on two TERT Biomarker investigation is a growing field in the man- mutations (c. 1-124C > T and c.1-146C > T) plus agement of NMIBC. Many of the investigations are used FGFR3 (p.R248C and p.S249C) hotspot mutations. in different scenarios: diagnosis, surveillance, and risk After a technical validation of the test, they achieved stratification of patients with NMIBC. Although many a73.5% of sensitivity and 93.2% of specificity. molecular marker tests have been developed to improve Springer et al. [26] have also analyzed mutations in diagnostic and surveillance accuracy, with some having TERT promoter, mutations in FGFR3 in combination been approved by the US Food and Drug Administra- with other nine gen mutations (TP53, CDKN2A, tion, none of the currently available tests have been ac- ERBB2, HRAS, KRAS, PIK3CA, MET, VHL, MLL) cepted or incorporated into the follow-up algorithms plus detection of aneuploidy,an abnormal chromo- described in the guidelines [51]. some number, that has been estimated to be present Biomarkers can be divided into cellular, protein and in > 90% of the cancer of most histopathologic types genetic markers. The latter are the most recent and, in [49]. They found that this combination could detect contrast to cytology, have the advantages of being reli- recurrences with a sensitivity of 68% and a specifi- able, easy to perform, and objective. city of 80%. In fact, they perform significantly better in BC because thousands of genetic changes can be detected accurately RNA tests (Table 2) and simultaneously compared with the lower- RNA biomarkers are less well studied in the field of BC throughput protein-based biomarkers. As aberrant DNA surveillance. methylation also occurs in non-malignant tissue it is not MicroRNAs (miRNAs) are B22-nucleotide long, pathognomonic of malignancy and genetic methylation single-stranded, non-coding RNAs that bind to comple- cannot be used to distinguish between cancer cells and mentary ‘seed’ regions found in the 30-untranslated re- other pre- or non-neoplastic cells [52]. However, this gion of particular messenger RNA (mRNA) species. genetic biomarker has the benefits of always occurring MiRNAs can modulate the expression of their mRNA in the same DNA location and chemical stability which targets and are involved in many physiological processes, make it easier to detect than gene mutations. but also in carcinogenesis [50]. Sapre et al. [29] evalu- Protein-based and cell-based biomarkers are also more ated a 12-miRNA-panel test, with the aim of selecting likely to be affected by benign conditions such as infec- the minimum number of miRNAs necessary to achieve tion, inflammation and bladder treatments. an accurate prediction. They found that a selection of To date the gold standard for these cases, as out- six miRNAs (miR16, miR200c, miR205, miR21, miR221 lined in the guidelines, is to use cystoscopy and cy- and miR34a) provided a sensitivity of 88% and a specifi- tology. Cystoscopy is an invasive procedure that may city of 48%. be associated with pain and discomfort [5]. More- Kavalieris et al. [30] and Lotan et al. [31] tested a com- over, cystoscopy does not detect all lesions and is bination of five mRNAs (IGFBP5, HOXA13, MDK, subject to the experience of the urologist or nurse CDK1, CXCR2), commercially available under the brand [53]. Voided urine cytology needs trained cytopathol- name Cx Bladder Monitor®, and reported highly consist- ogists and has the potential for inter-observer ent results for the evaluation of the mRNA expression variability. from the five genes. The studies included a scoring sys- Researchers who are developing urinary biomarkers tem, based on variables such as previous tumour status are looking for high sensitivity and a high negative pre- (primary or recurrent) and time since previous tumour dictive value. This profile is of special interest in the in years, to classify the test as positive or negative. The follow-up scenario because the aim of these tests is to authors reported sensitivities between 91 and 92% (95% reduce the number of cystoscopies by alternating the in high-grade tumours) and a negative predictive value procedures, rather than avoiding cystoscopy altogether. of 96%. Thus, cystoscopy will only be performed when the urine Wallace et al. [33] and Pichler et al. [32] tested the test is positive (urine-first strategy). Xpert BC Monitor®, a commercial kit that measures five One of the major limitations of the use of DNA- or target mRNAs (ABL1, CRH, IGF2, UPK1B, ANXA10), in mRNA-based techniques is the difficulty in obtaining a population of 510 patients and obtained sensitivities sufficiently large quantities of high-quality RNA from between 73 and 84% (100% in high-grade tumours) and voided urine. In terms of monitoring, another limitation a negative predictive value of 92–93%. They also of non-invasive urine biomarkers is their low sensitivity, Lozano et al. BMC Urology (2020) 20:99 Page 9 of 11 particularly for early-stage and low-grade tumours that sensitivity, especially for high-grade tumours, compared account for a significant proportion of recurrences. to the gold standard. European and US guidelines still Almost all the studies had a high percentage of ‘false’- recommend cystoscopy and cytology for follow-up. Gen- positive urine tests for the detection of concomitant etic urinary biomarkers are a very heterogeneous group recurrences, resulting in low specificity. In many articles, of test that nowadays cannot replace the standard path- the authors justified these percentages with the well- way of surveillance with cystoscopy and cytology. Al- known phenomenon of the anticipatory effect, i.e. the though there are some ongoing clinical trials comparing urine test detects recurrent tumours earlier than cystos- both options, there is no level 1 evidence to support copy. It is accepted in the literature that anticipatory their recommendation instead of the gold standard. detection would include recurrences that occur within Abbreviations the next 18 months after a positive biomarker test [54]. PRISMA: Preferred reporting items for systematic reviews and meta-analyses; In any case, performing a cystoscopy because of a false QUADAS: Quality Assessment of Diagnostic Accuracy Studies; BC: Bladder positive is more acceptable than missing a tumour be- cancer; EAU: European Association of Urology; AUA: American Urological Association; NMIBC: Non-muscle invasive bladder cancer; MA: Microsatellite cause of a false negative. analysis; TERT: Telomerase reverse transcriptase; MiRNA: MicroRNAs; Other limitations of the studies included in this review mRNA: Messenger RNA are the retrospective nature of some of the cohorts used Acknowledgements for the outcome analysis, artificial oversampling of the Not applicable. recurrence rate by recruiting patients scheduled for transurethral resection of a proven bladder tumour, and Authors’ contributions F. L and C.X.R. did the evidence acquisition and the tables. All authors (F.L., using the same population for the training and the valid- C.X.R., A.C., E.T., J.M.) contributed writing the manuscript, read and approved ation sets, which increases the possibility that the per- the final manuscript. formance of the biomarker may be artificially inflated Funding due to over-fitting. Not applicable. In this review, most of the biomarker tests are dichot- omous, providing either positive (tumour detected) or Availability of data and materials All data generated or analysed during this study are included in this negative (no tumour) test results. However, giving a nu- published article (and its supplementary information files). merical prediction of the probability of a recurrent tumour may be more helpful to urologists in terms of Ethics approval and consent to participate Not applicable. their decision-making. Moreover, there is a lack of uniformity in the design of Consent for publication the studies. Some of the works describe surveillance pro- Not applicable. grams but they create the cohorts. Many of the bio- Competing interests markers tested need clinical information to complete an The authors declare that they have no competing interests. algorithm and yield a positive or negative result, which Received: 27 February 2020 Accepted: 8 July 2020 increases subjectivity and decreases the homogeneity of results. The main limitations were the lack of randomized References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin. 2019; control trials and the diverse study outcomes, which 69(1):7–34. https://doi.org/10.3322/caac.21551. made meta-analysis impossible to perform. Comparison 2. Leal J, Luengo-Fernandez R, Sullivan R, Witjes JA. Economic burden of between sensitivity and specificity of different bio- bladder Cancer across the European Union. Eur Urol. 2016;69(3):438–47. https://doi.org/10.1016/j.eururo.2015.10.024. markers may generate a bias due to the different inci- 3. Compérat E, Gontero P, Mostafid AH, Palou J, Van Rhijn BWG, Rouprêt M, dence and different cohort. et al. Non-muscle-invasive Bladder Cancer (TaT1 and CIS) EAU Guidelines; Literature lacks of direct comparison between urinary 2018. p. 1–48. Retrieved from http://uroweb.org/wp-content/uploads/EAU- Guidelines-Non-muscle-invasive-Bladder-Cancer-TaT1-CIS-2018.pdf. biomarkers and gold standard maybe due to commercial 4. Daneshmand S, Konety BR. American urological association (AUA) guideline interests. American urological association non-muscle invasive bladder Cancer. AUA Clinical Guidelines, (April); 2016. p. 1–45. 5. Burke DM, Shackley DC, O’Reilly PH. The community-based morbidity of Conclusion flexible cystoscopy. BJU Int. 2002;89(4):347–9. https://doi.org/10.1046/j.1464- BC is one of the most expensive tumours due to its high 4096.2001.01899.x. recurrence rate and the costs of the follow-up protocols. 6. Karakiewicz PI, Benayoun S, Zippe C, Lüdecke G, Boman H, Sanchez-Carbayo M, et al. Institutional variability in the accuracy of urinary cytology for This is the reason why there is an increased interest in predicting recurrence of transitional cell carcinoma of the bladder. BJU Int. biomarkers, in order to reduce the number of explora- 2006;97(5):997–1001. https://doi.org/10.1111/j.1464-410X.2006.06036.x. tory investigations and improve the quality of life of pa- 7. Lotan Y, Roehrborn CG. Sensitivity and specificity of commonly available bladder tumor markers versus cytology: results of a comprehensive tients with BC. In this review, there are some genetic literature review and meta-analyses. Urology. 2003;61(1):109–18. https://doi. biomarkers with higher negative predictive value and org/10.1016/S0090-4295(02)02136-2. Lozano et al. BMC Urology (2020) 20:99 Page 10 of 11 8. Bensalah K, Montorsi F, Shariat SF. Challenges of Cancer biomarker profiling classifier for monitoring bladder cancer. Clin Epigenetics. 2018;10(1):1–10. {a figure is presented}. Eur Urol. 2007;52(6):1601–9. https://doi.org/10.1016/j. https://doi.org/10.1186/s13148-018-0496-x. eururo.2007.09.036. 25. Witjes JA, Morote J, Cornel EB, Gakis G, van Valenberg FJP, Lozano F, et al. 9. Kamat AM, Karakiewicz PI, Xylinas E, Hegarty PK, Hegarty N, Jenkins LC, et al. Performance of the bladder EpiCheck™ methylation test for patients under ICUD-EAU international consultation on bladder Cancer 2012: screening, surveillance for non–muscle-invasive bladder Cancer: results of a diagnosis, and molecular markers. Eur Urol. 2012;63(2013):4–15. multicenter, prospective, blinded clinical trial. Eur Urol Oncol. 2018;1(4):307– 10. Altman DG, McShane LM, Sauerbrei W, Taube SE, Cavenagh MM. REMARK 13. https://doi.org/10.1016/j.euo.2018.06.011. (REporting recommendations for tumor MARKer prognostic studies). 26. Springer SU, Chen CH, Del Carmen Rodriguez Pena M, Li L, Douville C, Guidelines for Reporting Health Research: A User’s Manual. 2014:241–9. Wang Y, et al. Non-invasive detection of urothelial cancer through the https://doi.org/10.1002/9781118715598.ch23. analysis of driver gene mutations and aneuploidy. ELife. 2018;7:1–27. https:// 11. Whiting, Penny F. ; Rutjes, Anne W.S.; Westwood, Marie E. ; Mallett, Susan ; doi.org/10.7554/eLife.32143. Deeks, Jonathan J. ; Reitsma, Johannes B., ; Leeflang, Mariska M.G. ; Sterne, 27. D’Andrea D, Soria F, Zehetmayer S, Gust KM, Korn S, Witjes JA, Shariat SF. Jonathan A.C. ; Bossuyt, P. M. M. ; (2011). QUADAS-2: a revised tool for the Diagnostic accuracy, clinical utility and influence on decision-making of a quality assessment of diagnostic accuracy studies. Ann Internal Med Res methylation urine biomarker test in the surveillance of non-muscle-invasive RepMethods, 155(4), 529–536. bladder cancer. BJU Int. 2019;123(6):959–67. https://doi.org/10.1111/bju.14673. 12. Campbell M, McKenzie JE, Sowden A, Katikireddi SV, Brennan SE, Ellis S, et al. 28. Batista R, Vinagre J, Prazeres H, Sampaio C, Peralta P, Conceição P, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting Validation of a novel, sensitive, and specific urine-based test for recurrence guideline. The BMJ. 2020;368:1–6. https://doi.org/10.1136/bmj.l6890. surveillance of patients with non-muscle-invasive bladder Cancer in a 13. Rouprêt M, Hupertan V, Yates DR, Comperat E, Catto JWF, Meuth M, et al. A comprehensive multicenter study. Front Genet. 2019;10(December):1–15. comparison of the performance of microsatellite and methylation urine https://doi.org/10.3389/fgene.2019.01237. analysis for predicting the recurrence of urothelial cell carcinoma, and 29. Sapre N, Macintyre G, Clarkson M, Naeem H, Cmero M, Kowalczyk A, et al. A definition of a set of markers by Bayesian network analysis. BJU Int. 2008; urinary microRNA signature can predict the presence of bladder urothelial 101(11):1448–53. https://doi.org/10.1111/j.1464-410X.2008.07591.x. carcinoma in patients undergoing surveillance. Br J Cancer. 2016;114(4):454– 14. van der Aa MNM, Zwarthoff EC, Steyerberg EW, Boogaard MW, Nijsen Y, van 62. https://doi.org/10.1038/bjc.2015.472. der Keur KA, et al. Microsatellite analysis of voided-urine samples for 30. Kavalieris L, O’Sullivan P, Frampton C, Guilford P, Darling D, Jacobson E, surveillance of low-grade non-muscle-invasive Urothelial carcinoma: et al. Performance characteristics of a multigene urine biomarker test for feasibility and clinical utility in a prospective multicenter study (cost- monitoring for recurrent Urothelial carcinoma in a multicenter study. J Urol. effectiveness of follow-up of urinary bladder Cancer trial C). Eur Urol. 2009; 2017;197(6):1419–26. https://doi.org/10.1016/j.juro.2016.12.010. 55(3):659–68. https://doi.org/10.1016/j.eururo.2008.05.001. 31. Lotan Y, OʼSullivan P, Raman JD, Shariat SF, Kavalieris L, Frampton C, et al. 15. Zuiverloon TCM, Van Der Aa MNM, Van Der Kwast TH, Steyerberg EW, Clinical comparison of noninvasive urine tests for ruling out recurrent Lingsma HF, Bangma CH, Zwarthoff EC. Fibroblast growth factor receptor 3 urothelial carcinoma. Urologic Oncol. 2017;35(8):531.e15–22. https://doi.org/ mutation analysis on voided urine for surveillance of patients with low- 10.1016/j.urolonc.2017.03.008. grade non-muscle - invasive bladder cancer. Clin Cancer Res. 2010;16(11): 32. Pichler R, Fritz J, Tulchiner G, Klinglmair G, Soleiman A, Horninger W, et al. 3011–8. https://doi.org/10.1158/1078-0432.CCR-09-3013. Increased accuracy of a novel mRNA-based urine test for bladder cancer 16. Reinert T, Borre M, Christiansen A, Hermann GG, Ørntoft TF, Dyrskjøt L. surveillance. BJU Int. 2018;121(1):29–37. https://doi.org/10.1111/bju.14019. Diagnosis of bladder Cancer recurrence based on urinary levels of EOMES, 33. Wallace E, Higuchi R, Satya M, McCann L, Sin MLY, Bridge JA, et al. HOXA9, POU4F2, TWIST1, VIM, and ZNF154 Hypermethylation. PLoS One. Development of a 90-minute integrated noninvasive urinary assay for 2012;7(10):1–9. https://doi.org/10.1371/journal.pone.0046297. bladder Cancer detection. J Urol. 2018;199(3):655–62. https://doi.org/10. 17. Zuiverloon TCM, Beukers W, Van Der Keur KA, Munoz JR, Bangma CH, 1016/j.juro.2017.09.141. Lingsma HF, et al. A methylation assay for the detection of non-muscle- 34. Mourah S, Cussenot O, Vimont V, Desgrandchamps F, Teillac P, Cochant- invasive bladder cancer (NMIBC) recurrences in voided urine. BJU Int. 2012; Priollet B, et al. Assessment of microsatellite instability in urine in the 109(6):941–8. https://doi.org/10.1111/j.1464-410X.2011.10428.x. detection of transitional-cell carcinoma of the bladder. Int J Cancer. 1998; 18. Allory Y, Beukers W, Sagrera A, Flández M, Marqués M, Márquez M, et al. 79(6):629–33. https://doi.org/10.1002/(SICI)1097-0215(19981218)79:6<629:: Telomerase reverse transcriptase promoter mutations in bladder cancer: AID-IJC13>3.0.CO;2-1. high frequency across stages, detection in urine, and lack of association 35. Van Oers JMM, Lurkin I, Van Exsel AJA, Nijsen Y, Van Rhijn BWG, Van Der Aa with outcome. Eur Urol. 2014;65(2):360–6. https://doi.org/10.1016/j.eururo. MNM, Zwarthoff EC. A simple and fast method for the simultaneous 2013.08.052. detection of nine fibroblast growth factor receptor 3 mutations in bladder 19. Abern MR, Owusu R, Inman BA. Clinical performance and utility of a DNA cancer and voided urine. Clin Cancer Res. 2005;11(21):7743–8. https://doi. methylation urine test for bladder cancer. Urologic Oncol. 2014;32(1):51. org/10.1158/1078-0432.CCR-05-1045. e21–6. https://doi.org/10.1016/j.urolonc.2013.08.003. 36. Heller G, Babinsky VN, Ziegler B, Weinzierl M, Noll C, Altenberger C, et al. 20. Su S-F, de Castro Abreu AL, Chihara Y, Tsai Y, Andreu-Vieyra C, Daneshmand Genome-wide CpG island methylation analyses in non-small cell lung S, et al. A panel of three markers hyper- and Hypomethylated in urine cancer patients. Carcinogenesis. 2013;34(3):513–21. https://doi.org/10.1093/ sediments accurately predicts bladder Cancer recurrence. Clin Cancer Res. carcin/bgs363. 2014;20(7):1978–89. https://doi.org/10.1158/1078-0432.CCR-13-2637. 37. Kim JG, Takeshima H, Niwa T, Rehnberg E, Shigematsu Y, Yoda Y, et al. 21. Fantony JJ, Abern MR, Gopalakrishna A, Owusu R, Jack Tay K, Lance RS, Comprehensive DNA methylation and extensive mutation analyses reveal Inman BA. Multi-institutional external validation of urinary TWIST1 and NID2 an association between the CpG island methylator phenotype and methylation as a diagnostic test for bladder cancer. Urologic Oncol. 2015; oncogenic mutations in gastric cancers. Cancer Lett. 2013;330(1):33–40. 33(9):387.e1–6. https://doi.org/10.1016/j.urolonc.2015.04.014. https://doi.org/10.1016/j.canlet.2012.11.022. 22. Beukers W, van der Keur KA, Kandimalla R, Vergouwe Y, Steyerberg EW, 38. Ying J, Li H, Seng TJ, Langford C, Srivastava G, Tsao SW, et al. Functional Boormans JL, et al. FGFR3, TERT and OTX1 as a urinary biomarker epigenetics identifies a protocadherin PCDH10 as a candidate tumor combination for surveillance of patients with bladder Cancer in a large suppressor for nasopharyngeal, esophageal and multiple other carcinomas prospective multicenter study. J Urol. 2017;197(6):1410–8. https://doi.org/10. with frequent methylation. Oncogene. 2006;25(7):1070–80. https://doi.org/ 1016/j.juro.2016.12.096. 10.1038/sj.onc.1209154. 23. Roperch JP, Grandchamp B, Desgrandchamps F, Mongiat-Artus P, Ravery V, 39. Kelly TK, Jones PA, Sharma S. Epigenetics in cancer. Carcinogenesis. 2009; Ouzaid I, et al. Promoter hypermethylation of HS3ST2, SEPTIN9 and SLIT2 31(1):27–36. https://doi.org/10.1093/carcin/bgp220. combined with FGFR3 mutations as a sensitive/specific urinary assay for 40. Saxonov S, Berg P, D. L. B. A genome-wide analysis of CpG dinucleotides in diagnosis and surveillance in patients with low or high-risk non-muscle- the human genome distinguishes two distinct classes of promoters. Proc invasive bladder cancer. BMC Cancer. 2016;16(1):1–9. https://doi.org/10. Natl Acad Sci. 2006;18(2):193–204. https://doi.org/10.1080/03071375.1994. 1186/s12885-016-2748-5. 9747015. 24. van der Heijden AG, Mengual L, Ingelmo-Torres M, Lozano JJ, van Rijt-van 41. Laird PW. The power and the promise of DNA methylation markers. Nat Rev de Westerlo CCM, Baixauli M, et al. Urine cell-based DNA methylation Cancer. 2003;3(4):253–66. https://doi.org/10.1038/nrc1045. Lozano et al. BMC Urology (2020) 20:99 Page 11 of 11 42. Reinert T, Modin C, Castano FM, Lamy P, Wojdacz TK, Hansen LL, et al. Comprehensive genome methylation analysis in bladder cancer: identification and validation of novel methylated genes and application of these as urinary tumor markers. Clin Cancer Res. 2011;17(17):5582–92. https://doi.org/10.1158/1078-0432.CCR-10-2659. 43. Serizawa RR, Ralfkiær U, Steven K, Lam GW, Schmiedel S, Schüz J, et al. Integrated genetic and epigenetic analysis of bladder cancer reveals an additive diagnostic value of FGFR3 mutations and hypermethylation events. Int J Cancer. 2011;129(1):78–87. https://doi.org/10.1002/ijc.25651. 44. Kandimalla R, Van Tilborg AAG, Kompier LC, Stumpel DJPM, Stam RW, Bangma CH, Zwarthoff EC. Genome-wide analysis of CpG Island methylation in bladder cancer identified TBX2, TBX3, GATA2, and ZIC4 as pTa-specific prognostic markers. Eur Urol. 2012;61(6):1245–56. https://doi.org/10.1016/j. eururo.2012.01.011. 45. Renard I, Joniau S, van Cleynenbreugel B, Collette C, Naômé C, Vlassenbroeck I, et al. Identification and validation of the methylated TWIST1 and NID2 genes through Real-time methylation-specific polymerase chain reaction assays for the noninvasive detection of primary bladder Cancer in urine samples. Eur Urol. 2010;58(1):96–104. https://doi.org/10.1016/j.eururo. 2009.07.041. 46. Cancer T, Atlas G. Comprehensive molecular characterization of Urothelial bladder carcinoma:the Cancer genome Atlas research network. Nature. 2013;507(7492):315–22. https://doi.org/10.1038/nature12965.Comprehensive. 47. Billerey C, Chopin D, Bralet M, Lahaye J, Abbou CC, Bonaventure J, et al. Short Communication. 2001;158(6):1955–9. 48. Killela PJ, Reitman ZJ, Jiao Y, Bettegowda C, Agrawal N, Diaz LA, et al. TERT promoter mutations occur frequently in gliomas and a subset of tumors derived from cells with low rates of self-renewal. Proc Natl Acad Sci U S A. 2013;110(15):6021–6. https://doi.org/10.1073/pnas.1303607110. 49. Douville C, Springer S, Kinde I, Cohen JD, Hruban RH, Lennon AM, et al. Detection of aneuploidy in patients with cancer through amplification of long interspersed nucleotide elements (LINEs). Proc Natl Acad Sci U S A. 2018;115(8):1871–6. https://doi.org/10.1073/pnas.1717846115. 50. Kiselev FL. MicroRNA and cancer. Mol Biol. 2014;48(2):232–42. 51. Lotan Y, Shariat SF, Schmitz-Dräger BJ, Sanchez-Carbayo M, Jankevicius F, Racioppi M, et al. Considerations on implementing diagnostic markers into clinical decision making in bladder cancer. Urologic Oncol. 2010;28(4):441–8. https://doi.org/10.1016/j.urolonc.2009.11.004. 52. Esteller M. CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future. Oncogene. 2002;21(35 REV. ISS. 3):5427– 40. https://doi.org/10.1038/sj.onc.1205600. 53. van der Aa MNM, Steyerberg EW, Bangma C, van Rhijn BWG, Zwarthoff EC, van der Kwast TH. Cystoscopy revisited as the gold standard for detecting bladder Cancer recurrence: diagnostic review Bias in the randomized, prospective CEFUB trial. J Urol. 2010;183(1):76–80. https://doi.org/10.1016/j. juro.2009.08.150. 54. Wolff EM, Chihara Y, Pan F, Weisenberger DJ, Siegmund KD, Sugano K, et al. Unique DNA methylation patterns distinguish noninvasive and invasive urothelial cancers and establish an epigenetic field defect in premalignant tissue. Cancer Res. 2010;70(20):8169–78. https://doi.org/10.1158/0008-5472. CAN-10-1335. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.