Urologic Oncology: Seminars and Original Investigations ] (]]]]) ]]]–]]]

Seminar article

Considerations on the use of urine markers in the management of patients with high-grade non–muscle-invasive bladder cancer Ashish M. Kamat, M.D.a, Antonia Vlahou, Ph.D.b, John A. Taylor, M.D.c, M´Liss A. Hudson, M.D.d, Beate Pesch, Ph.D.e, Molly A. Ingersoll, Ph.D.f, Tilmann Todenhöfer, M.D.g, Bas van Rhijn, M.D.h, Wassim Kassouf, M.D.i, H. Barton Grossman, M.D.a, Thomas Behrens, Ph.D.j, Ashish Chandra, M.D.k, Peter J. Goebell, M.D.l, Juan Palou, M.D.m, Marta Sanchez-Carbayo, Ph.D.n, Bernd J. Schmitz-Dräger, M.D., Ph.D.o,* a

Department of Urology, University of Texas MD Anderson Cancer Center, Houston, TX Division of Biotechnology, Biomedical Research Foundation, Academy of Athens, Athens, Greece c Division of Urology, University of Connecticut Health Center, Farmington, CT d Ochsner Clinic Foundation, Tom and Gayle Benson Cancer Center, New Orleans, LA e Institute for Prevention and Occupational Medicine of the German Social Accident Insurance (IPA), Institute of the Ruhr University, Bochum, Germany f Department of Immunology, Institut Pasteur and U818 Inserm, Paris, France g Urologische Universitätsklinik, Eberhard Karls Universität, Tübingen, Germany h Division of Surgical Oncology (Urology), Antoni van Leeuwenhoek Hospital, Netherlands Cancer Institute, Amsterdam, The Netherlands i McGill University Health Center, Montreal, Canada j Protein Research Unit Ruhr within Europe (PURE), Ruhr-Universität, Bochum, Germany k Cellular Pathology, St Thomas' Hospital, London, UK l Urologische Klinik, Friedrich-Alexander-Universität, Erlangen, Germany m Servicio de Urología, Fundación Puigvert, Barcelona, Spain n CIC bioGUNE, Bilbao, Spain o Urologie Schön Klinik Nürnberg Fürth/Urologie, Fürth, Germany b

Abstract Objective: Diagnosis and surveillance of high risk non muscle-invasive bladder cancer (NMIBC) represent specific challenges to urologists. In contrast to low/intermediate risk tumors, these tumors recur more frequently. A significant number will eventually progress to muscleinvasive bladder cancer, a life threatening disease requiring extensive therapeutic efforts. Although clinical risk factors have been identified that may predict tumor recurrence and progression, additional biomarkers are desperately needed to improve tumor diagnosis and guide clinical management of these patients. In this article, the role of molecular urine markers in the management of high risk NMIBC is analyzed. Methods: In this context, several potential indications (diagnostic, prognostic, predictive) were identified and the requirements for molecular markers were defined. In addition, current knowledge within the different indications was summarized. Results: Significant progress has been made in the last decade studying the impact of molecular urine markers in patients with high risk NMIBC. Conclusions: Although we may not be ready for the inclusion of molecular markers in clinical decision-making, and many questions remain unanswered, recent studies have identified situations in which the use of molecular markers in particular in high grade tumors may prove beneficial for patient diagnosis and surveillance. r 2014 Elsevier Inc. All rights reserved. Keywords: Urine markers; Non–muscle-invasive bladder cancer; High grade; Diagnosis; Disease management

This article reflects and summarizes discussions held at the 10th Meeting of the International Bladder Cancer Network (IBCN e.V.), Nijmegen, The Netherlands, 20—22.9.2012. * Corresponding author. Tel.: þ49-911-9714-531; fax: þ49-911-9714532. E-mail address: [email protected] (B.J. Schmitz-Dräger). http://dx.doi.org/10.1016/j.urolonc.2014.06.017 1078-1439/r 2014 Elsevier Inc. All rights reserved.

Introduction High-risk non–muscle-invasive bladder cancer (NMIBC) is characterized by tumors that recur frequently and often progress to muscle-invasive bladder cancer, a deadly disease.

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

2

Table 1 FDA-approved urine tests for diagnosis or surveillance of patients with bladder cancer or both Assay

Source

Target

Assay type

Urine cytology BTA stat (Polymedco) BTA TRAK (Polymedco) NMP22 (Alere) NMP22 BladderChek (Alere) ImmunoCyt/uCytþ (Scimedx)

Cells Urine Urine Urine Urine Cells

Morphology Complement factor H and complement factor H-related protein Complement factor H and complement factor H-related protein Nuclear matrix protein 22 Nuclear matrix protein 22 Two mucin glycoproteins, high molecular carcinoembryonic antigen, and morphology Morphology, alterations in chromosomes 3, 7, 17, and 9p21

Microscopy Immunoassay or point-of-care device Sandwich ELISA Sandwich ELISA point-of-care device Immunofluorescence microscopy

UroVysion (Abbott, Vysis)

FISH

ELISA ¼ enzyme-linked immunosorbent assay.

Thus, urologists are faced with the specific challenges of early diagnosis and lifelong patient surveillance. To reduce tumor recurrence rates and potentially halt tumor progression, reresection of the tumor bed and maintenance intravesical BCG instillation are recommended to patients with high-risk NMIBC. Several clinical risk factors have been identified that may predict tumor recurrence and progression in NMIBC [reviewed in Refs. 1 and 2]; however, additional biomarkers would likely improve tumor diagnosis and guide clinical management of these patients. Through the last decades, numerous molecular urine markers for diagnosis of bladder cancer (BC) have been developed. Although some of them have been approved for diagnostic use and surveillance by the Food and Drug Administration (FDA) (Table 1), no molecular marker has been incorporated into guideline recommendations or clinical decision making for patients with high-risk NMIBC. It may be argued that the performance of current markers may

be not sufficient, but multiple studies and reviews demonstrate that most diagnostic urine markers are more sensitive than conventional urine cytology, including in high-grade disease [3–6] (Tables 2 and 3). We therefore conclude that marker evaluation has been insufficient in the past, which is exemplified by the fact that in contrast to low/intermediaterisk BC [2], randomized controlled trials comparing the use of urine markers vs. standard care, including cystoscopy and urine cytology, are still lacking in high-risk NMIBC. There is a growing body of evidence that molecular markers may predict tumor progression and risk-stratify patients who are being treated with intravesical therapies. In particular, cell cycle–regulating genes, epigenetic events (e.g., altered methylation), and apoptotic genes have been suggested to contribute to defining the prognosis of patients with BC [1,7,8,9]. Using markers such as fluorescence in situ hybridization (FISH) to identify patients who are likely to fail intravesical therapy would allow urologists the

Table 2 Marker sensitivity and specificity of cytology and commercially available markers (data from reviews/meta-analyses) Marker

Median sensitivity (range)

Median specificity (range)

Total number of patients

Cytology Lotan and Roehrborn [3] van Rhijn et al. [4] Mowatt et al. [5]

34 (2053) 35 (1375) 44 (3851)a

99 (8399) 94 (85100) 96 (9498)a

2,767 5,545 14,260

BTA stat Lotan and Roehrborn [3] van Rhijn et al. [4]

71 (5782) 58 (2974)

73 (6182) 73 (5686)

2,534 3,461

NMP22 (assay/BladderChek) Lotan and Roehrborn [3] van Rhijn et al. [4] Mowatt et al. [5] (pooled) Mowatt et al. [5] (BladderChek)

73 71 68 65

80 73 79 81

uCytþ/Immunocyt van Rhijn et al. [4] Mowatt et al. [5] Schmitz-Dräger et al. [6] FISH (Urovysion) Mowatt et al. [5] Schmitz-Dräger et al. [6] a

95% CI.

(4787) (47100) (6274)a (5085)

(5891) (5598) (7484)a (4087)

2,413 2,041 10,119 2,426

67 (52100) 84 (7791)a 81 (42100)

75 (6282) 75 (6883)a 75 (6295)

959 3,041 4,899

76 (6584)a 72 (23100)

85 (7892)a 80 (40100)

3,101 2,852

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

3

Table 3 Sensitivity of cytology and commercially available markers (data from reviews/meta-analyses) based on tumor grade Marker

No. of studies

Grade 1

Grade 2

Grade 3

Cytology Lotan and Roehrborn [3] van Rhijn et al. [4]

8 9

0.12 (0.04–0.31) 0.17

0.26 (0.17–0.37) 0.34

0.64 (0.38–0.84) 0.58

BTA stat Lotan and Roehrborn [3] van Rhijn et al. [4]

8 7

0.47 (0.38–0.56) 0.45

0.73 (0.59–0.83) 0.6

0.94 (0.55–0.99) 0.75

NMP22 van Rhijn et al. [4] Lotan and Roehrborn [3]

3 7

0.41 0.61 (0.35–0.81)

0.53 0.71 (0.41–0.90)

0.8 0.79 (0.63–0.89)

Immunocyt van Rhijn et al. [4] Schmitz-Dräger et al. [6]

1 19

0.78 0.75

0.9 0.84

1 0.84

FISH (Urovysion) van Rhijn et al. [4] Schmitz-Dräger et al. [6]

2 21

0.56 0.53

0.78 0.81

0.95 0.79

possibility to offer patients an alternative therapy/radical cystectomy in a timely manner, before tumor progression [10,11]. It is the mission of the International Bladder Cancer Network to assess the usefulness of molecular markers in clinical practice. Although another article in this series investigates the role of molecular urine markers in the diagnosis of low/intermediate BC [2], here we have analyzed the current molecular markers available to support diagnosis and prediction of response to treatment in patients with high-risk BC. High-risk NMIBC, as defined by the European Association of Urology guidelines, is any T1 tumor; any high-grade or grade 3 tumor; carcinoma in situ (CIS); or multiple, recurrent, and large (43 cm) Ta grade 1 or grade 2 tumors [12]. According to the risk tables developed by the European Organization for Research and Therapy of Cancer (EORTC) [13] (Tables 4), high-risk patients are characterized by a progression score of 46, which translates into a 5-year progression rate of more than 8%. For the purposes of this discussion, we limit our comments to primary or recurrent high-grade (World Health Organization 2004 grading system)/grade 3 (World Health Organization 1973 grading system) pTa, pT1 tumors and CIS [14,15]. In this article, the following questions are discussed:

 How can we incorporate urinary molecular markers into  

clinical decision making in the management of highrisk NMIBC? What are the requirements for urine markers in the management of patients with high-risk NMIBC? What is the current status of diagnostic molecular urine markers in diagnosis/surveillance of patients with highrisk NMIBC?

In this concept paper, we primarily focus on the FDAapproved markers (Table 1). This decision was not made because the authors believe that these markers are superior but only based on the fact that these markers are better studied than other markers. Furthermore, these markers are commercially available to the urologic community and therefore can be easily included in clinical studies and, eventually, in clinical routine.

How can we incorporate urinary molecular markers into clinical decision making in the management of high-risk NMIBC? Molecular markers can be applied to improve diagnosis of tumor recurrence, guide the physician, and help counsel patients for radical vs. conservative treatment (prognosis), or predict response to therapy. Assessment of molecular BC markers includes the examination of gene or protein expression in urine, blood, or tissue. Although prognostic molecular markers are primarily tissue based for BC diagnosis, urine sediment or supernatant is rarely used. Various scenarios wherein urine makers may potentially guide physicians in clinical decision making in patients with high-grade NMIBC include the following: 1. Patient surveillance: Unlike follow-up for low/intermediate-risk NMIBC, where delayed tumor detection is unlikely to cause harm, the sensitivity of a urinary marker in highgrade tumors must be very high to replace cystoscopy. At present, there is no molecular marker with sufficiently high sensitivity to reduce the frequency or entirely replace cystoscopy in patients with high-risk disease. Thus, urine markers may be used as an adjunct to cystoscopy in a

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

4

Table 4 EORTC risk tables for risk assessment of tumor recurrence and progression in NMIBC [13] Factor Number of tumours Single 2–7 Z Tumor diameter o3 cm Z3 cm Prior recurrence rate Primary r1 recurrence/y 41 recurrence/y Category Ta T1 Concurrent CIS No Yes Grade (WHO 1973) G1 G2 G3 Total score Recurrence score

Recurrence

Progression

0 3 6

0 3 3

0 3

0 3

0 2 4

0 2 2

0 1

0 4

0 1

0 6

0 1 2 017

0 0 5 023

3.

Probability of recurrence Probability of recurrence at 1 y at 5 y %

(95% CI)

%

(95% CI)

0 14 59 1017

15 24 38 61

(1019) (2126) (3541) (5567)

31 46 62 78

(2437) (4249) (5865) (7384)

Progression score

Probability of progression at 1 y

Probability of progression at 5 y

%

(95% CI)

%

(95% CI)

0.2 1 5 17

(00.7) (0.41.6) (47) (1024)

0.8 6 17 45

(01.7) (58) (1420) (3555)

0 26 713 1423

2.

4.

5.

WHO ¼ World Health Organization.

high-risk BC scenario. Indeed, conventional follow-up using office cystoscopy and urine cytology may miss tumor recurrence in a significant number of cases [16–19]. Molecular markers could potentially synergize with the performance of conventional surveillance, improving overall sensitivity. Knowing the results of a molecular marker test before performing diagnostic procedures such as cystoscopy would likely improve tumor detection. This is particularly desirable when managing patients with highrisk disease, where early detection is critical. Reflex testing may be another option to use urine markers for improving BC surveillance. The addition of a sensitive molecular marker in patients with negative findings of urine cytology may improve the sensitivity of conventional follow-up. In a study including 505 patients undergoing surveillance for NMIBC, the performance of

6.

FISH in patients with negative findings of cytology led to a sensitivity of 79.8% (94.0% for high-grade, T1, or CIS lesions) [20]. Generating awareness (positive molecular marker result with negative clinical workup): As the specificity of urine cytology is high, positive findings of cytology combined with normal findings of upper tract imaging and a negative finding of cystoscopy are an indication for further investigation with random bladder and prostatic biopsies to rule out CIS. If findings of voided urine cytology are positive and those of random bladder/ prostatic biopsies are negative, additional examination may be added in this setting to ensure that high-grade/ CIS lesions are identified early. Hexaminolevulinate (fluorescence cystoscopy) has been shown to enhance the detection of CIS when compared with white light cystoscopy and can be used to perform targeted biopsies of abnormal lesions [21]. If results of all investigations are negative, ureterorenoscopy and random biopsies of the upper tract may be considered. In general, urine markers have the potential to outperform urine cytology in this setting owing to their higher sensitivity. Identifying at-risk patients: Detecting patients at risk for tumor progression may be another challenge for (prognostic) molecular markers. This type of research is usually done with tissue-based molecular analyses; however, urine-based approaches can be envisioned. Identifying BCG responsiveness: Molecular markers could be used to predict patient outcome after intravesical therapy. These markers can be either urine or tissue based. Prediction of response to BCG treatment is of particular relevance in high-risk NMIBC to avoid treatment-related side effects and unnecessary costs. The ability to predict nonresponsiveness might allow urologists to consider radical surgery before metastases develop. Predicting tumor aggressiveness: Molecular markers to predict the aggressiveness of the tumor before performing transurethral resection (TUR) may influence the decision to use hexaminolevulinate photodynamic resection or include removal of additional random biopsies. Surveillance of patients with urinary diversions: Although controversial, some groups have integrated surveillance of the urethra and upper urinary tract after radical cystectomy to detect secondary urinary tract tumors early. This may be particularly useful in patients with high risk of extravesical urinary tract recurrence such as patients with urethral or ureteral tumors/CIS in cystectomy specimens or patients with diffuse BC or protracted history of NMIBC failing intravesical therapy.

What are the requirements for urine markers in management of patients with high-risk NMIBC? To integrate a new marker into clinical decision making, it must be superior to standard measures and variables

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

currently in use. Superiority may comprise better performance/accuracy, more rapid test results (point of care), and decreased costs. Evidence to support the use of a marker would preferably be derived from prospective studies with a large sample size in a multicenter setting. In addition, validation of the marker in a randomized fashion would be desirable. It should be noted that phase III trials in marker research are exceptionally rare and very difficult to conduct. This has led to the Oxford Levels of Evidence Committee to accept well-designed cohort studies as grade I Levels of Evidence [22]. When discussing suitable molecular markers to support the diagnosis and surveillance of high-grade BC, it becomes obvious that sensitivity will be of much higher importance as compared with low/intermediate-risk BC. Although the most significant side effect of a false-positive finding is an unnecessary endoscopy or upper urinary tract assessment, a false-negative result could have disastrous consequences. Based on these considerations, assays with a high sensitivity for high-grade tumors are needed. In this context, it is important to consider that the reported assay sensitivity is lower in patients with BC who are under surveillance as compared with those at primary diagnosis. This observation has been linked to the smaller size of recurrent tumors. Boman et al. [23] found that NMP22 has a higher sensitivity for primary tumors as compared with recurrent tumors, because of higher stage and grade at presentation and larger tumor size. For 187 patients evaluated with NMP22 (cutoff 414.6 U/ml), Sánchez-Carbayo et al. [24] reported sensitivities of 83% (19/23), 81% (34/42), and 93% (38/41) in detecting tumors of o5, 5 to 30, and 430 mm, respectively. The same observations were reported in a systematic review on the use of urine markers for NMIBC surveillance [4]. For commercially available and widely used assays, most tests exceed the sensitivity reported for urine cytology (Table 2) [3–6]. When comparing the sensitivities of urine cytology and molecular marker tests for high-grade BC detection, the gap narrows. However, the sensitivities of commercially available molecular markers are still approximately 15% to 20% higher than that of urine cytology (Table 3). Although specificity of urine cytology is higher than that of molecular urine markers, this difference may not be of utmost importance for surveillance of patients with high-grade NMIBC. Based upon these considerations, molecular markers may be more useful than urine cytology in detecting high-grade BC. Follow-up of patients undergoing BCG therapy represents a specific challenge for molecular urine assays, as inflammatory products in urine have been shown to compromise test results [25]. Ongoing BCG treatment and urinary tract infection are contraindications for some assays. Of the FDA-approved markers, the NMP22 assays (NMP22 BladderChek and NMP22 Bladder Cancer Test) do not appear to be influenced by BCG administration, even though other confounding variables exist, including comorbidities of the urogenital tract (kidney stones, infections, etc.) [27].

5

Patients with BC who have urinary diversion may not be candidates for use of molecular marker tests for surveillance because their urine contains nonurothelial cells, which cross-react with agents in some molecular assays. Despite limited experience and poor data quality, it appears that assays using genetic changes or epitopes also present in mucosal cells (e.g., mucins and carcinoembryonic antigens) are not suitable for patients with BC who have urinary diversions.

What is the current status of diagnostic molecular urine markers in diagnosis/surveillance of patients with high-risk NMIBC? Several molecular markers are currently under evaluation for use in surveillance of high-grade NMIBC [reviewed in Refs. 5, 6, and 26]. The NMP22 BladderChek was assessed in a prospective study of 2222 patients with a history of NMIBC [27]. In 581 patients diagnosed with BC, NMP22 provided good sensitivity (50%–90% depending on cancer stage), was associated with recurrence and progression, and improved cancer detection in high-grade patients. Nevertheless, the effect of the test results such as these on clinical decision making in the management of high-grade NMIBC remains unclear [28]. FISH was recently evaluated for surveillance of patients undergoing BCG intravesical therapy [10]. In this prospective study of 126 patients, samples were analyzed before and following initiation of BCG treatment (at 6, 12, and 24 wk). Positive test results were associated with higher tumor recurrence and progression rates, suggesting that FISH holds potential in monitoring treatment efficacy as well as predicting tumor progression. A recent meta-analysis of FISH demonstrated that overall performance was superior to urine cytology (area under the curve: 87% vs. 63%) [29]. This difference, however, was almost entirely attributable to the difference in performance in diagnosing patients with pTa lesions; improved performance disappeared when patients with pTa lesions were excluded from the analysis (area under the curve: 94% vs. 91%). A prognostic value of FGFR3 mutations in T1 tumors has also been suggested [1,38]. In this study of 132 patients with incident pT1 BC, an FGFR3 mutation (detected in 28% of cases) was associated with a favorable prognosis; a validation study is currently underway. As FGFR3 mutations can also be detected in the urine, a positive test result would not only be highly specific for tumor recurrence but also permit the identification of patients at low risk of developing tumor progression. The only published phase III study including marker results for clinical decision making investigated microsatellite analysis (MA) in urine samples in surveillance of 448 patients with NMIBC [16,30]. A sensitivity of 70% was observed in the intervention arm where the urologist was aware of the MA result as compared with 29% in the

6

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

control arm of the study (urologist unaware of the MA result). This study represents a milestone in marker research. However, this study included patients with low/ intermediate-risk NMIBC, and high-risk NMIBC was one of the exclusion criteria. Further randomized markers studies are urgently needed for the integration of urine markers into clinical decision making. As cell-based assays are subject to interobserver variation, direct comparisons between cell-based assays and molecular marker tests are difficult to interpret. Data from meta-analyses suggest that cell-based assays (e.g., FISH and immunocytology) may be more sensitive if compared to assays studying soluble compounds in urine. However, this difference is higher in low-grade tumors but less pronounced in high-grade disease [2,5,6]. This observation may be explained by the fact that current cell-based assays also include markers for low-grade tumors, whereas urinebased assays mostly examine a single protein. Furthermore, cell-based markers or genetic markers may be less confounded by urinary tract infection than protein-based markers as the target proteins may be differently active. Given the number of tests available and their varied specificities, one strategy may be to specifically combine markers from different molecular “levels” [16,31,32]. However, multimarker testing harbors the risk of an impaired specificity, leading to overdiagnosis and unnecessary concerns for the patient. Although there is a relatively high rate of false-positive results translating into a relatively low positive predictive value of the FISH test, findings from several studies suggest that the low specificity in follow-up trials may be explained in part by anticipatory positive results in cases where premalignant changes precede the cystoscopic discovery of recurrent malignancy [33–35]. Yafi et al. [17] reported that in patients with positive urine cytology and a complete negative workup, urothelial carcinoma is eventually detected in 44% of patients during further follow-up. Similarly, anticipation of a positive result with FISH analysis in the setting of a negative workup has been demonstrated to predict subsequent bladder tumors in up to 63% of patients [36]. Other studies found that 85% to 89% of the patients with a “false-positive” test eventually developed tumor recurrence [34,35]. The role that anticipation of a positive result plays in surveillance of patients is unclear as many patients with NMIBC eventually experience disease recurrence whereas others stay free of disease. It is obvious that there must be a trade-off between early diagnosis and false-positive findings. Pesch et al. analyzed chromosomal instability (CIN) using FISH in 200,000 single cells from samples in the UroScreen trial [33]. As early as 2 to 4 years before diagnosis, an increased number of cells with CIN were detected but the extent of CIN in these samples was lower than the cutoff value for a positive test. Nonetheless, anticipating positive results likely generates an awareness of for potential risk, which would be beneficial for patients

with high-risk BC. A potential strategy to explore this question is to design a prospective trial in which patients with a positive result of a urine test are treated with intravesical instillations and repeated urine marker testing. Multiple other markers are currently under evaluation, including tests to assess DNA alterations, RNA and protein expression levels, and epigenetic changes. In a prospective study of 42 patients with intermediate- to high-risk NMIBC receiving BCG treatment, the prognostic relevance of cell cycle markers, including p53, p21, waf1/cip, Bcl-2, cyclinD1, and metallothionein 9, was analyzed in TUR tissue samples [38]. Association of these markers individually, or in combination with EORTC risk scores, with disease outcome (recurrence, progression, and survival) was evaluated. The results supported an added value of Bcl-2, cyclinD1, metallothionein, and p21 to the current EORTC prognosticators in predicting disease recurrence or progression or both and underscore the need for further validation studies. Urine markers may help to predict the aggressiveness of a lesion observed in surveillance cystoscopy. In a recent study, combined application of cytology, immunocytology, FISH, and NMP22 improved risk stratification before TUR of the bladder was performed and helped to identify highrisk tumors [39]. This may be relevant, as the current guidelines recommend blue light TUR only in patients suspected of having a high-grade tumor. The performance of urine markers in the follow-up of patients with urinary diversion after radical cystectomy for BC needs further evaluation. In a recent retrospective analysis of 275 patients following radical cystectomy, the specificity of urine cytology (86%) and FISH (87%) was considered unsatisfactory [40].

Conclusions Although it may be argued that the consideration of the use of diagnostic urine markers cannot separately address low-grade and high-grade disease, there are several arguments supporting a separate consideration: the fact that most studies discriminate between different tumor grades when reporting sensitivity/specificity suggests the relevance of this approach. The reason underlying is that sensitivity of most markers correlates with tumor grade. Furthermore, it is obvious that we look at different tumor entities from the molecular point of view, which also translates into a different biological behavior. In consequence, the clinical expectations to diagnostic markers vary, whereas in lowgrade tumors, detection of tumor recurrence is the primary end point and a delayed detection is not dangerous [2]. In high-risk disease, highly sensitive assays (even with anticipatory properties) are required to not oversee tumor recurrence and persistence (CIS) or disease progression. In consequence, although for low-grade tumors reduction of cystoscopies may be envisioned, in high-grade/high-risk disease we attempt to improve sensitivity of endoscopy.

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

Effective clinical management of high-grade NMIBC remains challenging with key issues being a timely diagnosis of disease recurrence or progression or both. Timely diagnosis is necessary to prevent progression to muscleinvasive cancer while avoiding unnecessary procedures in the case of less aggressive BC. If muscle-invasive tumor is not found at rebiopsy, continuous follow-up cystoscopy and urinary cytology are recommended along with annual upper urinary tract imaging. Nevertheless, a significant number of tumors will recur and eventually progress to muscleinvasive disease. Current surveillance modalities are imperfect in providing prompt and timely detection of recurrences or progression to muscle-invasive disease and thus, negatively [9,19,41,42], affect patient survival. van der Aa et al. [16] have demonstrated the feasibility of including MA into clinical decision making. For patients with high-risk BC, improved detection of tumor recurrence is essential for cancer control and may eventually prove to be life saving. Further validation of this concept using commercially available assays is necessary. Similar to low/ intermediate-risk BC, the question remains as to whether single marker testing is sufficiently sensitive or whether multiple marker tests may be superior [2]. Systematic reviews clearly point at a positive correlation between sensitivity of diagnostic tumor markers (urine cytology) and tumor grade (Table 3). For most commercial assays a sensitivity between 80% and 95% for grade 3 tumors has been reported [3–6]. An increased expression of tumor-associated antigens in high-grade disease is the most likely reason for this observation. Tumor load, mostly larger in high-grade disease, may be another factor improving sensitivity. Based upon these considerations, the high sensitivity required for diagnosis of high-grade tumors may be achieved by using 1 or 2 assays. In consequence, the role of urine cytology in high-grade urothelial cancer, where sensitivity is of paramount relevance, requires reconsideration. The issue of anticipation of positive results is severely understudied. Although technical difficulties in studying this problem are acknowledged, more data would be helpful in better understanding its usefulness in predicting future cancer occurrence. In particular, genetic alterations (e.g., FISH and MA) appear to be promising markers for investigation of the role of this particular aspect of patient surveillance. Improved risk substratification is needed for T1 cancers, which are currently associated with the highest risk of overtreatment or undertreatment [26]. Substaging classification systems for T1 tumors have been proposed based on penetration of the muscularis mucosae, measured depth in mm from mucosal border, or extent of tumor invading the lamina propria (e.g., T1a, T1b, T1c, or more recently T1-microinvasive and T1-extensive invasive). None of these classifications has been adopted in clinical practice, as pathologists have not reached a consensus agreement on their clinical usefulness/reproducibility. Biomarkers providing improved risk stratification could have a major effect on

7

clinical management by facilitating treatment selection for the individual patient. Determination of FGFR3 mutations may be 1 factor that could play an important role in determining the outcome to different treatment modalities for patients with T1 tumors [37]. The frequency and duration of the surveillance program for cancer-free BCG-treated patients remain unclear. Cases of recurrence even after 10 years are not infrequent. If reliable prediction of treatment response through noninvasive biomarker testing could be achieved, this would likely increase patient compliance with follow-up procedures and would reduce costs. In addition, nonresponders may be offered alternative treatment modalities in time to avoid disease progression. Cost-benefit issues are not addressed in this review, as we feel that the time for this type of consideration has not yet arrived. In a first step, suitable assays need to be selected and their performance must be demonstrated in a randomized controlled study. The data obtained in this type of trial will then provide a solid ground for cost-benefit calculations. Summary Significant progress has been made during the last decade studying the effect of molecular urine markers in patients with high-grade urothelial cancer. Still, key questions remain unanswered. For example, the management of patients with persistent abnormal markers in the face of normal endoscopy, biopsy, and imaging is not yet well defined, and systematic analysis of this problem is missing. Nevertheless, although inclusion of molecular markers into clinical decision has not been rigorously validated yet, recent studies have identified scenarios in which the use of markers may prove to be beneficial. References [1] van Rhijn BWG, Catto JW, Goebell PJ, Knüchel R, Shariat SF, Sanchez-Carbayo M, et al. Molecular markers for urothelial bladder cancer prognosis: towards implementation in clinical practice. Urol Oncol [this issue]. [2] Schmitz-Dräger BJ, Todenhöfer T, van Rhijn B, Pesch B, Hudson MA, Chandra A, et al. Considerations on the use of urine markers in the management of patients with low/intermediate risk non muscleinvasive bladder cancer. Urol Oncol 2014:[pii:S1078-1439(13) 00450-X]. [3] Lotan Y, Roehrborn CG. Sensitivity and specificity of commonly available bladder tumor markers versus cytology: results of a comprehensive literature review and meta-analyses. Urology 2003;61:109–18:[discussion 118]. [4] van Rhijn BWG, van der Poel HG, van der Kwast TH. Urine markers for bladder cancer surveillance: a systematic review. Eur Urol 2005;47:736–48. [5] Mowatt G, Zhu S, Kilonzo M, et al. Systematic review of the clinical effectiveness and cost-effectiveness of photodynamic diagnosis and urine biomarkers (FISH, ImmunoCyt, NMP22) and cytology for the detection and follow-up of bladder cancer. Health Technol Assess 2010;14:1–331:[iii-iv].

8

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9

[6] Schmitz-Dräger BJ, Droller M, Lotan Y, Lokeshwar VB, van Rhijn B, Hemstreet GP, et al. Molecular markers for bladder cancer screening, early diagnosis and surveillance. In: Soloway M, Khoury S. editors Bladder cancer: 2nd international consultation on bladder tumors, vol 21. Paris, France: Editions; 2012. p. 171-205. [7] Goebell PJ, Groshen S, Schmitz-Dräger BJ, for the INTERNATIONAL STUDY-INITIATIVE ON BLADDER CANCER (ISBC): p53 immunohistochemistry in bladder cancer—a new approach to an old question. Urol Oncol 2010;28:377–88. [8] Shariat SF, Droller M, Fradet Y, Hemstreet G, Hudson MA, Karakiewicz P, et al. Molecular markers for bladder cancer staging and prognosis. In: Soloway M, Khoury S. editors Bladder cancer: 2nd international consultation on bladder tumors. vol 21. Paris, France: Editions; 2012. p. 207–29. [9] Kamat AM, Hegarty PK, Gee JR, Clark PE, Svatek RS, Hegarty N, et al. ICUD-EAU International Consultation on Bladder Cancer 2012: screening, diagnosis, and molecular markers. Eur Urol 2013;63:4–15. [10] Kamat AM, Dickstein RJ, Messetti F, Anderson R, Pretzsch SM, Gonzalez GN, et al. Use of fluorescence in situ hybridization to predict response to bacillus Calmette-Guerin therapy for bladder cancer: results of a prospective trial. J Urol 2012;187:862–7. [11] Zuiverloon TC, Nieuweboer AJ, Vékony H, Kirkels WJ, Bangma CH, Zwarthoff EC. Markers predicting response to bacillus CalmetteGuérin immunotherapy in high-risk bladder cancer patients: a systematic review. Eur Urol 2012;61:128–45. [12] Babjuk M, Burger M, Zigeuner R, Shariat S, van Rhijn B, Compérat E, et al. Guidelines on non–muscle invasive bladder cancer Ta, T1 and CIS. European Association of Urology; 2013. Available at: http:// www.uroweb.org/gls/pdf/05_TaT1_Bladder_Cancer_LR.pdf. [13] Sylvester RJ, van der Meijden AP, Oosterlinck W, Witjes JA, Bouffioux C, Denis L, et al. Predicting recurrence and progression in individual patients with stage TaT1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006;49:466–75. [14] Epstein JI, Amin MB, Reuter VR, et al. The World Health Organization/International Society of Urological Pathology consensus classification of urothelial (transitional cell) neoplasms of the urinary bladder. Am J Surg Pathol 1998;22:1435–48. [15] Sauter G, Algaba F, Amin M, et al. Tumours of the urinary system: non-invasive urothelial neoplasias. In: Eble JN, Sauter G, Jl Epstein, Sesterhenn I, editors. WHO classification of classification of tumours of the urinary system and male genital organs. Lyon: IARCC Press; 2004. p. 29–34. [16] van der Aa MN, Steyerberg EW, Bangma C, van Rhijn BW, Zwarthoff EC, van der Kwast TH. Cystoscopy revisited as the gold standard for detection of bladder cancer recurrences: diagnostic review bias in a randomised prospective trial (Cefub-Trial). J Urol 2010;183:76–80. [17] Yafi FA, Brimo F, Auger M, Aprikian A, Tanguay S, Kassouf W. Is the performance of urinary cytology as high as reported historically? A contemporary analysis in the detection and surveillance of bladder cancer. Urol Oncol 2013:[pii:S1078-1439(12)00343-2]. [18] Schmitz-Dräger BJ, Tirsar L-A, Schmitz-Dräger C, Dörsam J, Mellan Z, Bismarck E, et al. Immunocytology in the assessment of patients with asymptomatic hematuria. World J Urol 2008;26:31–7. [19] Messer J, Shariat SF, Brien JC, Herman MP, Ng CK, Scherr DS, et al. Urinary cytology has a poor performance for predicting invasive or high-grade upper-tract urothelial carcinoma. BJU Int 2011;108:701–5. [20] Todenhöfer T, Hennenlotter J, Esser M, Aufderklamm S, Mohrhardt S, Gakis G, et al. Stepwise applications of immunocytology and fish in the surveillance of patients with non-muscle invasive bladder cancer. J Urol 2013;189S:e526. [21] Burger M, Grossman HB, Droller M, Schmidbauer J, Hermann G, Drăgoescu O, et al. Photodynamic diagnosis of non–muscle-invasive bladder cancer with hexaminolevulinate cystoscopy: a meta-analysis of detection and recurrence based on raw data. Eur Urol 2013;64:846–54.

[22] Goebell PJ, Kamat AM, Sylvester R, Black P, Droller M, Godoy G, et al. Assessing the quality of studies on the diagnostic accuracy of tumor markers. Urol Oncol [this issue]. [23] Boman H, Hedelin H, Holmäng S. Four bladder tumor markers have a disappointingly low sensitivity for small size and low grade recurrence. J Urol 2002;167:80–3. [24] Sánchez-Carbayo M, Herrero E, Megías J, Mira A, Soria F. Comparative sensitivity of urinary CYFRA 21-1, urinary bladder cancer antigen, tissue polypeptide antigen, tissue polypeptide antigen and NMP22 to detect bladder cancer. J Urol 1999;162:1951–6. [25] Raitanen MP, FinnBladder Group: The role of BTA stat test in follow-up of patients with bladder cancer: results from FinnBladder studies. World J Urol 2008;26:45–50. [26] van Rhijn BW. Combining molecular and pathologic data to prognosticate non–muscle-invasive bladder cancer. Urol Oncol 2012;30: 518–23. [27] Shariat SF, Savage C, Chromecki TF, Sun M, Scherr DS, Lee RK, et al. Assessing the clinical benefit of nuclear matrix protein 22 in the surveillance of patients with nonmuscle-invasive bladder cancer and negative cytology: a decision-curve analysis. Cancer 2011;117: 2892–7. [28] Barbieri CE, Cha EK, Chromecki TF, Dunning A, Lotan Y, Svatek RS, et al. Decision curve analysis assessing the clinical benefit of NMP22 in the detection of bladder cancer: secondary analysis of a prospective trial. BJU Int 2012;109:685–90. [29] Hajdinjak T. UroVysion FISH test for detecting urothelial cancers: meta-analysis of diagnostic accuracy and comparison with urinary cytology testing. Urol Oncol 2008;26:646–51. [30] van der Aa MN, Zwarthoff EC, Steyerberg EW, Boogaard MW, Nijsen Y, van der Keur KA, et al. Microsatellite analysis of voidedurine samples for surveillance of low-grade non–muscle-invasive urothelial carcinoma: feasibility and clinical utility in a prospective multicenter study (Cost-Effectiveness of Follow-Up of Urinary Bladder Cancer Trial [CEFUB]). Eur Urol 2009;55:659–68. [31] Todenhöfer T, Hennenlotter J, Esser M, Mohrhardt S, Tews V, Aufderklamm S, et al. Combined application of cytology and molecular urine markers to improve the detection of urothelial carcinoma. Cancer Cytopathol 2013;121:252–60. [32] Pesch B, Brüning T, Johnen G, Casjens S, Bonberg N, Taeger D, et al. Biomarker research with prospective study designs for the early detection of cancer. Biochim Biophys Acta 2014;1844(5): 874–83. [33] Yoder BJ, Skacel M, Hedgepeth R, Babineau D, Ulchaker JC, Liou LS, et al. Reflex UroVysion testing of bladder cancer surveillance patients with equivocal or negative urine cytology: a prospective study with focus on the natural history of anticipatory positive findings. Am J Clin Pathol 2007;127:295–301. [34] Gofrit ON, Zorn KC, Silvestre J, Shalhav AL, Zagaja GP, Msezane LP, et al. The predictive value of multi-targeted fluorescent in-situ hybridization in patients with history of bladder cancer. Urol Oncol 2008;26:246–9. [35] Skacel M, Fahmy M, Brainard JA, Pettay JD, Biscotti CV, Liou LS, et al. Multitarget fluorescence in situ hybridization assay detects transitional cell carcinoma in the majority of patients with bladder cancer and atypical or negative urine cytology. J Urol 2003;169:2101–5. [36] Lotan Y, Bensalah K, Ruddell T, Shariat SF, Sagalowsky AI, Ashfaq R. Prospective evaluation of the clinical usefulness of reflex fluorescence in situ hybridization assay in patients with atypical cytology for the detection of urothelial carcinoma of the bladder. J Urol 2008;179: 2164–9. [37] van Rhijn BW, van der Kwast TH, Liu L, Fleshner NE, Bostrom PJ, Vis AN, et al. The FGFR3 mutation is related to favorable pT1 bladder cancer. J Urol 2012;187:310–4. [38] Alkhateeb SS, Neill M, Bar-Moshe S, Rhijn BV, Kakiashvili DM, Fleshner N, et al. Long-term prognostic value of the combination of EORTC risk group calculator and molecular markers in non–muscle-invasive bladder

A.M. Kamat et al. / Urologic Oncology: Seminars and Original Investigations ] (]]]]) 1–9 cancer patients treated with intravesical Bacille Calmette-Guérin. Urol Ann 2011;3:119–26. [39] Todenhöfer T, Hennenlotter J, Aufderklamm S, Kühs U, Gakis G, Germann M, et al. Individual risk assessment in bladder cancer patients based on a multi-marker panel. J Cancer Res Clin Oncol 2013;139:49–56. [40] Fernández MI, Parikh S, Grossman HB, Katz R, Matin SF, Dinney CP, et al. The role of FISH and cytology in upper urinary tract surveillance

9

after radical cystectomy for bladder cancer. Urol Oncol 2012;30: 821–4. [41] Wallace DM, Bryan RT, Dunn JA, Begum G, Bathers S, West Midlands Urological Research Group: Delay and survival in bladder cancer. BJU Int 2002;89:868–78. [42] Hollenbeck BK, Dunn RL, Ye Z, Hollingsworth JM, Skolarus TA, Kim SP, et al. Delays in diagnosis and bladder cancer mortality. Cancer 2010;116:5235–42.