World J Urol (2008) 26:511

DOI 10.1007/s00345-008-0240-9

TOPIC PAPER

Guidelines for development of diagnostic markers

in bladder cancer
Peter J. Goebell Susan L. Groshen
Bernd J. Schmitz-Drger

Received: 3 October 2007 / Accepted: 19 January 2008 / Published online: 6 February 2008
Springer-Verlag 2008

Abstract Conclusion The implementation of this discussion may

Objectives Advances in understanding cancer at the foster the integration of new tools and strategies screening,
molecular level have identiWed numerous alterations asso- diagnosis and surveillance of patients with bladder cancer.
ciated with cancer development and progression. The
eVorts in evaluating these putative biomarkers in clinical Keywords Marker Bladder Cancer Phases
studies of patients with cancer are increasing in order to
improve the clinical management of the disease. However,
despite numerous attempts, the results of such biomarker Introduction
studies are frequently inconsistent and sometimes even
contradictory. Aim of this work is to discuss some of the Advances in understanding cancer at the molecular level
recognized problems which have impeded our understand- have identiWed numerous alterations associated with cancer
ing of the role of new markers and prevented the introduc- development and progression. In case of bladder cancer,
tion of these markers into patient management. there is a strong need for new markers supporting screen-
Results These Problems include standardization issues, ing, initial diagnosis, surveillance for recurrent lesions,
selection of patient cohorts and endpoints and statistical detection of early progression, and prediction of the biolog-
considerations. In order to improve and standardize marker ical potential of a particular tumor with the ultimate aim to
development a stepwise procedure in four phases, analo- alter clinical patient management [18].
gous to clinical trials is proposed. Furthermore, a common With voided urine there is a unique source available for
terminology, considerations on the population to study as bladder cancer diagnosis. This medium can be collected
well as general recommendations for planning and conduct- absolutely non-invasive and at more or less unlimited quan-
ing the evaluation of markers will be presented. tity. In consequence, it appears relatively easy to develop a
bladder cancer test, since many molecular determinants of
bladder tumor growth and invasion are released into urine
[9, 10]. Thus, many bladder cancer tests, such as BTA-Stat/
TRAK, UBC-Rapid, UBC-IRMA, BLCA-4, HA-HAase,
P. J. Goebell (&)
Department of Urology, NMP-22, and survivin detect soluble markers released in
Friedrich-Alexander University of Erlangen, urine (product inserts of BTA-Stat/TRAK, NMP-22, UBC)
Maximiliansplatz 1, 91054 Erlangen, Germany and are used to evaluate their clinical use [1124].
e-mail: peter.goebell@uk-erlangen.de
Moreover, due to a reduced cell-cell adherence prefera-
S. L. Groshen bly tumor cells are released into urine. This fact has not
Department of Preventive Medicine Norris Comprehensive only been the rationale for the development of urine cytol-
Cancer Center, University of Southern California, ogy but also other cell-based tests such as UroVysion (mul-
Keck School of Medicine, Los Angeles, CA, USA
ticolor FISH), uCyt+, microsatellite DNA analysis,
B. J. Schmitz-Drger telomerase, DD23, etc., detect either genetic alterations or
Department of Urology, EuromedClinic, Frth, Germany cell-surface antigens that may indicate biologic characteris-

123
6 World J Urol (2008) 26:511

tics of a particular cancer and may be used for treatment develop and assess biomarkers for diagnostic and prognos-
decisions [2540]. tic use, was and is to take a closer look at general methodo-
In addition, a wide range of alterations has been logical issues associated with Marker development and
detected through high-throughput analyses of tissue sam- evaluation [63]. The IBCN reached consensus, that in
ples comparing malignant lesions with adjacent normal establishing the utility of a marker for clinical use, investi-
tissue [4147]. However, there is a substantial lack of com- gators must demonstrate that (a) the marker can be reliably
prehensive studies investigating the presence of the identi- and consistently measured, (b) the marker has good sensi-
Wed molecular alterations from high throughput analyses in tivity and speciWcityso it can, with reasonably high prob-
serum or urine samples from the same individual. Serum ability, identify patients with a better or worse prognosis or
levels of a given alteration may diVer and be less informa- with a greater or lesser likelihood of having a speciWed con-
tive or proteins apparently important for the ability of a ditionand (c) the use of the marker will improve outcome
tumor cell to spread may not be found in urine. by targeting therapeutic or diagnostic interventions.
In summary, numerous attempts have been made to pro- Accordingly, this group has deWned four phases through
vide clinicians and consequentially patients with such a which markers are developed [64].
marker-test. Many of these markers have been extensively
studied in case-control trials. However, the results of these
studies are often inconsistent and sometimes contradictory. Phases of marker studies
Recognized problems include diVerent methods of per-
forming assays, the use of diVerent subsets of patients Phase I: assay development and evaluation of clinical
(diVerence in stage or treatment) and endpoints (e.g. local prevalence (feasibility studies)
versus distant recurrence versus survival), and inadequate
study design, leading to incompatible data sets. This has This phase comprises the identiWcation of a target poten-
impeded understanding the role of new markers. Since tially suited for diagnostic use. IdentiWcation of the target
replication and independent conWrmation are hallmarks of may occur on many ways, classically by identifying the tar-
the scientiWc method, the implementation of standards and get in tumor cells, however, with the advent of molecular
conventions is essential prior to any conclusive recom- technology other ways or deWnitions of a variety of target
mendation for the introduction of new markers to the are conceivable. Key issue remains that a diVerence
clinic [48, 49]. New markers are typically evaluated in sin- between tumor cells and normal urothelial cells is prerequi-
gle institutional settings and comparison due to diVerences site and must be demonstrated. It has to be of note that Weld
in study design, experimental methods, and data analysis eVects are an established integral part of the development
remains diYcult [50, 51]. Another often underestimated of bladder cancer, which warrants not only including nor-
obstacle is the fact, that there is no common terminology mal adjacent tissue but also tissue and samples from
in place for authors to describe the actual position of a healthy individuals as important controls.
given biomarker in its development. Thus, the circum- Furthermore, an adequate medium to be studied must be
stances and purpose under which a tumor-biologic deter- identiWed. For bladder cancer this will be urine is most
minant is evaluated and eventually used in a clinical cases, however, other sources are conceivable. A reproduc-
setting should be well deWned and clearly indicated, since ible and optimized assay is the essential prerequisite prior
its level of evidence may hamper its clinical use [52, 53]. to the application to clinical samples.
Currently, it is diYcult to conclude from reports whether a This should be complemented with feasibility studies
determinant could be useful for risk assessment (who may documenting the prevalence and expression of the markers
get cancer?), diagnosis (who has cancer?), the prediction of interest and examining the association with demographic
of outcome (prognosis), the prediction of response to a and clinical characteristics in a representative study cohort
given therapy (prediction) or as a possible target for inter- (target population). A case-control setting will be suYcient
vention. in this state of development.
Although there has been discussion for establishing gen-
eral methodological principles and guidelines for design, Phase II: evaluation studies for clinical utility
conduct, analysis, and reporting of marker studies (analo-
gous to those for clinical trials) [48, 5462], they are not This phase comprises further optimization of the assay
generally accepted. Therefore, funding agencies and technique (e.g., standardization, automatisation) and/or
research groups will need standardized guidelines for the interpretation of the assay results develop. The ultimate
conduction of maker studies in the future. One of the tasks goal of this phase of investigation is to develop hypotheses
of the International Bladder Cancer Network (IBCN), a and to deWne standards that can be used to perform the
multi-institutional interdisciplinary consortium formed to Phase III studies.

123
World J Urol (2008) 26:511 7

One of the reasons that no marker has been included in of cystoscopies in low risk patients or additional cystoscop-
clinical guidelines until today is a lack of adequate assess- ies or even biopsies in high risk patients with a positive test.
ment in a distinct clinical setting. Therefore, in this phase it If applicable, the new test must be compared with the
is necessary to deWne indications of future marker use and current standard (currently cytology in most situations).
to prospectively assess a given marker in this scenario. For SpeciWc attention must be paid to the fact that side-by-side
bladder cancer markers evaluation in the following indica- comparision must consider the nature of the assay. There-
tions is required: fore, it must be guaranteed that a suYcient expertise for all
tests included in the study is available. This is of speciWc
Screening
relevance for cell-based assays (e.g. urine cytology, FISH,
Assessment of patients with hematuria (gross/microhe-
immunocytology). Otherwise, the results for these tests
maturia)
might be negatively biased. If available, quality control
Follow-up of patients with bladder cancer
measures must be taken.
The reason behind is that a diagnosis of bladder cancer An elaborate study protocol also deWning the cut-oV val-
per se does not exist as a clinical question. Patients require ues prospectively and approval by an ethics committee as
examination due to, mostly unspeciWc symptoms or are well as sample size calculation are mandatory; a random-
investigated for tumor recurrence after a previous episode ized, multi-center study design is routineously applied.
of bladder cancer. Screening of high risk populations may
be another potential indication [65], however, it should be Phase IV: validation and technology transfer
noted that screening studies for bladder cancer are mostly as application studies
carried out prospectively in a large, well deWned cohort
based upon an elaborate protocol and thus may rather be The aims of Phase IV studies are (a) to transfer the tech-
considered phase III. Ethical approval should be obtained niques and established methods of the assays and other
although at this stage no consequences are drawn from the aspects of the technology and (b) to evaluate the ability of
results. other investigators and clinicians at other institutions to
Phase II trials are mostly mono-institutional studies. apply these methods and interpret the results.
However, adequately sized and representative samples of In consequence, standardization issues and quality con-
patients may be easier to achieve in a large collaborative trol measures need to be addressed. If the assay is to be per-
network with suYcient numbers of specimens to deWne and formed in centralized labs, sample procurement, storage
select the most appropriate set of samples. In addition, and shipping need to be deWned. In addition, cost/beneWt
identifying sources of variability during these phases of analyses may be done based upon information obtained
biomarker development is required for designing a Phase through phase III and/or phase IV trials.
III study. The Phase IV study is the Wnal step in the translational
Based upon the results adequate cut-oV values will be research process, in which a given biomarker is incorpo-
deWned for quantitative assays. It is essential that the out- rated into clinical practice. Phase IV trials are prospective
come from Phase II studies translates into hypotheses that multi-center studies. The establishment of a collaborative
may form the basis for Phase III analysis. network would clearly facilitate these essential interac-
tions.
Phase III: conWrmation studies

In Phase III hypotheses generated previously are tested Which population would beneWt ?
with suYcient power in a larger deWned clinical setting in
an independent, prospective cohort of patients. The clinical Cystoscopy, the gold standard for the detection of bladder
utility of a given marker assay, its performance, and inter- cancer, is invasive and relatively expensive. In fact, due to
pretation is established in that phase. The aim of Phase III the lifelong need for recurrence monitoring by cystoscopy
trials is the generation of information (evidence-based) that and the treatment of recurrent tumors, the cost per bladder
may eventually be included into clinical guidelines. cancer patient from diagnosis to death is the highest among
In order to comply with this goal it is necessary to all cancers [65].
include the marker into decision making in this phase and Thus, a noninvasive, highly sensitive and speciWc
to upfront deWne consequences drawn from a given result marker for detecting bladder cancer could decrease the
(positive or negative). E.g. for screening studies this would morbidity associated with cystoscopy, improve patient
result in urological assessment of individuals that tested quality of life, anddepending on the assay costdecrease
positive for a marker. For follow-up analyses after bladder costs by substituting a less expensive, noninvasive test for
cancer a negative test could lead to a decreased frequency the more expensive endoscopic procedure.

123
8 World J Urol (2008) 26:511

In consequence, it has to be discussed whether the intro- The potential of non-invasive markers in the assessment
duction of a biologic determinant/marker is useful in spe- of patients with hematuria was recently demonstrated by
ciWc a setting (Table 1). several authors [7477]. SpeciWcally in the assessment of
The majority of patients with high grade bladder cancer, patients with gross hematuria sensitive assays might
at the time of initial presentation, have tumors that invade improve diagnosis of bladder cancer in this population.
the lamina propria (stage T1) and beyond [8, 15, 6668].
These patients have a high risk of developing distant metas-
tasis and cancer speciWc death despite aggressive treatment. Material and datamandatory steps to harmonize
Early detection of bladder tumors, before they become the procurement of samples
muscleinvasive, might improve survival [69, 70]. Insofar,
screening of a high risk population could be of strong inter- As indicated above, the proper collection of samples will be
est [65, 70]. a crucial step in the entire process and needs a precise, reli-
Bladder tumors recur frequently; typically 4080% of able and transparent description of the used standard oper-
patients with bladder cancer will have a recurrence within ating procedures. In addition, the annotaion of the clinical
3 years following initial treatment [68]. Furthermore, data linked to the sample should also follow common rules
tumors that recur may be of higher grade or stage. Thus, to allow comparison of the results in the future. The harmo-
patients with bladder cancer usually undergo 3- to 6-month nization of biorepositories used for marker research and the
surveillance. To avoid unneccessary cystoscopies and/or to implementation of a common infrastructure as in clinical
have a non-invasive tool for the decision on whom to per- study-networks and interdisciplinary multi-institutional
form a cystoscopy would clearly be beneWcial. Thus, for the groups will facilitate the exchance for meaningful research
surveillance of Bladder Cancer the investigation of biologic in the Weld as it has for other tumor enteties [7880].
determinants are urgently needed and would deWnite alter
the management of the disease [7073].
Statistical considerations

Table 1 The use of biologic determinants in diVerent clinical settings Statistical assessment of marker trials may vary considering
Indication Potential role of markers the phase of development of a given marker or the targeted
study population.
Screening (q)
Phase 1 studies will make for example use of gene-
Microhematuria (q)
expression analyses and yield in hundreds of altered species
Gross hematuria q
of mRNA. The reliability and reproducability may be cru-
Voiding symptoms - cial not to obscure important determinants. The ability of
Follow up q the analyses to discriminate between a cancerous or non-

Table 2 Possible work-Xow for a Phase I study

To do

1. DeWne Primary aims

IdentiWcation of leads for potentially useful Review previous work on follow-up schedules
determinants for follow-up and crucial steps during the monitoring of a patient
Prioritize identiWed leads Identify and prioritize beneWcial role in the schedule for a marker
2. Select specimen
Consider/determine variability of subjects in demographics, Review of variability in the follow-up population
histology, prognosis, mode of detection
Consider/determine inXuence of disease prevalence Review prevalence of primary aims
3. Primary outcome measures
Value of biologic determinant Identify measures that control for assay variability
4. Evaluating results
Binary scale Provide true-positive and false-positive rates to determine
sensitivity and speciWcity
Multiple values Provide Receiver operating curve (ROC)
Development of a statistical algorithm for the use of the determinant
5. Plan conWrmatory study

123
World J Urol (2008) 26:511 9

cancerous lesion as measured on a binary scale as well as 11. Glas AS, Roos D, Deutekom M, Zwinderman AH, Bossuyt PM,
true positive and true negative values will be indicative for Kurth KH (2003) Tumor markers in the diagnosis of primary blad-
der cancer. A systematic review. J Urol 169:19751982
the robustness of the results. In addition, sensitivity and 12. Babjuk M, Kostrov M, Mudra K, Pecher S, Smolov H, Pecen L,
speciWcity will be applicable as reciever operating curves Ibrahim Z, Dvorcek J, Jarolm L, Novk J, Zima T (2002)
(ROC) as descriptive parameters of the test used in case the Qualitative and quantitative detection of urinary human comple-
alteration may take many values being more or less indica- ment factor H-related protein (BTA stat and BTA TRAK) and
fragments of cytokeratins 8, 18 (UBC rapid and UBC IRMA) as
tive for malignant chances [8183]. markers for transitional cell carcinoma of the bladder. Eur Urol
To exemplify these methodological considerations a 41:3439
possible work-Xow for a Phase I study to identify determi- 13. Gibanel R, Ribal MJ, Filella X, Ballesta AM, Molina R, Alcaraz
nants during follow-up is provided in Table 2 (modiWed A, Alcover JB (2002) BTA TRAK urine test increases the eYcacy
of cytology in the diagnosis of low-grade transitional cell carci-
acc. to [84]). noma of the bladder. Anticancer Res 22:11571160
14. Serretta V, Pomara G, Rizzo I, Esposito E (2000) Urinary BTA-
stat, BTA-trak and NMP22 in surveillance after TUR of recurrent
Conclusions superWcial transitional cell carcinoma of the bladder. Eur Urol
38:419425
15. Lokeshwar VB, Soloway MS (2001) Current bladder tumor tests:
There is a clear need to further harmonize terminology, does their projected utility fulWll clinical necessity? J Urol
approaches and circumstances under which markers are 165:10671077
evaluated in the future. The successful transition of marker/ 16. Chiong E, Gaston KE, Grossman HB (2007) Urinary markers in
screening patients with hematuria. World J Urol [Epub ahead of
determinants into useful clinical tools to either detect ini- print]
viduals at risk or to avoid unnecessary invasive procedures 17. Nguyen CT, Jones JS (2007) DeWning the role of NMP22 in blad-
will depend on the implementation of general methodologi- der cancer surveillance. World J Urol [Epub ahead of print]
cal principles and guidelines for design, conduct, analysis 18. Black PC, Brown GA, Dinney CP (2006) Molecular markers of
urothelial cancer and their use in the monitoring of superWcial uro-
and reporting of marker studies. thelial cancer. J Clin Oncol 24:55285535
19. Tritschler S, Scharf S, Karl A, Tilki D, Knuechel R, Hartmann A,
ConXict of interest statement There is no conXict of interest. Stief CG, Zaak D (2007) Validation of the diagnostic value of
NMP22 BladderChek test as a marker for bladder cancer by pho-
todynamic diagnosis. Eur Urol 51:403407
20. Shariat SF, Marberger MJ, Lotan Y, Sanchez-Carbayo M, Zippe
References C, Ldecke G, Boman H, Sawczuk I, Friedrich MG, Casella R,
Mian C, Eissa S, Akaza H, Serretta V, Huland H, Hedelin H, Raina
1. Malkin A (1992) Tumor markers. In: Tannock IF, Hill RP (eds) R, Miyanaga N, Sagalowsky AI, Roehrborn CG, Karakiewicz PI
The basic science of oncology, 2nd edn. McGraw-Hill, New York (2006) Variability in the performance of nuclear matrix protein 22
2. Gudjnsson S, Isfoss BL, Hansson K, Domanski AM, Warenholt J, for the detection of bladder cancer. J Urol 176:919926
Soller W, Lundberg LM, Liedberg F, Grabe M, Mnsson W (2007) 21. Moonen PM, Kiemeney LA, Witjes JA (2005) Urinary NMP22
The value of the urovysion((R)) Assay for surveillance of non-mus- BladderChek test in the diagnosis of superWcial bladder cancer.
cle-invasive bladder cancer. Eur Urol [Epub ahead of print] Eur Urol 48:951956
3. Birkhahn M, Mitra AP, Cote RJ (2007) Molecular markers for 22. Shariat SF, Zippe C, Ldecke G, Boman H, Sanchez-Carbayo M,
bladder cancer: the road to a multimarker approach. Expert Rev Casella R, Mian C, Friedrich MG, Eissa S, Akaza H, Sawczuk I,
Anticancer Ther 7:17171727 Serretta V, Huland H, Hedelin H, Rupesh R, Miyanaga N,
4. Herrmann E, Eltze E, Bierer S, Kpke T, Grge T, Neumann J, Sagalowsky AI, Wians F Jr, Roehrborn CG, Lotan Y, Perrotte P,
Hertle L, WlWng C (2007) VEGF-C, VEGF-D and Flt-4 in tran- Benayoun S, Marberger MJ, Karakiewicz PI (2005) Nomograms
sitional bladder cancer: relationships to clinicopathological including nuclear matrix protein 22 for prediction of disease recur-
parameters and long-term survival. Anticancer Res 27:31273133 rence and progression in patients with Ta, T1 or CIS transitional
5. Tsai YS, Tzai TS, Chow NH (2007) Does HER2 immunoreactivity cell carcinoma of the bladder. J Urol 173:15181525
provide prognostic information in locally advanced urothelial car- 23. Chang YH, Wu CH, Lee YL, Huang PH, Kao YL, Shiau MY
cinoma patients receiving adjuvant M-VEC chemotherapy? Urol (2004) Evaluation of nuclear matrix protein-22 as a clinical diag-
Int 79:210216 nostic marker for bladder cancer. Urology 64:687692
6. Knowles MA (2007) Role of FGFR3 in urothelial cell carcinoma: 24. May M, Hakenberg OW, Gunia S, Pohling P, Helke C, Lbbe L,
biomarker and potential therapeutic target. World J Urol 25:581 Nowack R, Siegsmund M, Hoschke B (2007) Comparative
593 diagnostic value of urine cytology, UBC-ELISA, and Xuorescence
7. Mohammed A, Khan Z, Zamora I, Bhatti A (2008) Biological in situ hybridization for detection of transitional cell carcinoma of
markers in the diagnosis of recurrent bladder cancer: an overview. urinary bladder in routine clinical practice. Urology 70:449453
Expert Rev Mol Diagn 8:6372 25. Friedrich MG, Toma MI, Hellstern A, Pantel K, Weisenberger DJ,
8. Droller MJ (2001) Cancer heterogeneity and its biologic implica- Noldus J, Huland H (2003) Comparison of multitarget Xuores-
tions in the grading of urothelial carcinoma. J Urol 165:696697 cence in situ hybridization in urine with other noninvasive tests for
9. Lokeshwar VB, Civantos F (2004) Tumor markers: current status. detecting bladder cancer. BJU Int 92:911914
In: Droller MJ (eds) American cancer society atlas of clinical 26. Toma MI, Friedrich MG, Hautmann SH, Jkel KT, Erbersdobler
oncology: bladder cancer. Hamilton, Ontario, pp 160205 A, Hellstern A, Huland H (2004) Comparison of the ImmunoCyt
10. Konety BR, Getzenberg RH (2001) Urine based markers of test and urinary cytology with other urine tests in the detection and
urological malignancy. J Urol 165:600611 surveillance of bladder cancer. World J Urol 22:145149

123
10 World J Urol (2008) 26:511

27. Greene KL, Berry A, Konety BR (2006) Diagnostic Utility of the ers: a high throughput tissue microarray analysis. BJU Int
ImmunoCyt/uCyt+ Test in Bladder Cancer. Rev Urol Fall 8:190 100:11821187
197 44. Dyrskjot L, Zieger K, Orntoft TF (2007) Recent advances in high-
28. Schmitz-Drger BJ, Tirsar LA, Schmitz-Drger C, Drsam J, Me- throughput molecular marker identiWcation for superWcial and
llan Z, Bismarck E, Ebert T (2007) Immunocytology in the assess- invasive bladder cancers. Front Biosci 12:20632073
ment of patients with asymptomatic hematuria. World J Urol 45. Wild PJ, Herr A, Wissmann C, Stoehr R, Rosenthal A, Zaak D, Si-
[Epub ahead of print] mon R, Knuechel R, Pilarsky C, Hartmann A (2005) Gene expres-
29. Mian C, Maier K, Comploj E, Lodde M, Berner L, Lusuardi L, Pal- sion proWling of progressive papillary noninvasive carcinomas of
ermo S, Vittadello F, Pycha A (2006) uCyt+/ImmunoCyt in the the urinary bladder. Clin Cancer Res 11:44154429
detection of recurrent urothelial carcinoma: an update on 1991 46. Snchez-Carbayo M, Cordon-Cardo C (2003) Applications of ar-
analyses. Cancer 108:6065 ray technology: identiWcation of molecular targets in bladder can-
30. Lodde M, Mian C, Comploj E, Palermo S, Longhi E, Marberger cer. Br J Cancer 89:21722177
M, Pycha A (2006) uCyt+ test: alternative to cystoscopy for less- 47. Zaharieva BM, Simon R, Diener PA, Ackermann D, Maurer R, Al-
invasive follow-up of patients with low risk of urothelial carci- und G, Knnagel H, Rist M, Wilber K, Hering F, Schnenberger
noma. Urology 67:950954 A, Flury R, Jger P, Fehr JL, Mihatsch MJ, Gasser T, Sauter G,
31. Szarvas T, Kovalszky I, Bedi K, Szendroi A, Majoros A, Riesz P, Toncheva DI (2003) High-throughput tissue microarray analysis
Fle T, Lszl V, Kiss A, Romics I (2007) Deletion analysis of tu- of 11q13 gene ampliWcation (CCND1, FGF3, FGF4, EMS1) in
mor and urinary DNA to detect bladder cancer: urine supernatant urinary bladder cancer. J Pathol 201:603608
versus urine sediment. Oncol Rep 18:405409 48. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M,
32. Frigerio S, Padberg BC, Strebel RT, Lenggenhager DM, Messthal- Clark GM (2006) Statistics Subcommittee of NCI-EORTC Work-
er A, Abdou MT, Moch H, Zimmermann DR (2007) Improved ing Group on Cancer Diagnostics. Reporting recommendations for
detection of bladder carcinoma cells in voided urine by standard- tumor MARKer prognostic studies (REMARK). Breast Cancer
ized microsatellite analysis. Int J Cancer 121:329338 Res Treat 100:229235
33. Burger M, Burger SJ, Denzinger S, Wild PJ, Wieland WF, Blaszyk 49. Real FX (2007) p53: it has it all, but will it make it to the clinic as
H, Obermann EC, Stoehr R, Hartmann A (2006) Elevated micro- a marker in bladder cancer? J Clin Oncol 25:53415344
satellite instability at selected tetranucleotide repeats does not cor- 50. Kyzas PA, Denaxa-Kyza D, Ioannidis JP (2007) Almost all arti-
relate with clinicopathologic features of bladder cancer. Eur Urol cles on cancer prognostic markers report statistically signiWcant
50:770775 results. Eur J Cancer 43:25592579
34. Fornari D, Steven K, Hansen AB, Jepsen JV, Poulsen AL, Vibits 51. Kyzas PA, Denaxa-Kyza D, Ioannidis JP (2007) Quality of report-
H, Horn T (2006) Transitional cell bladder tumor: predicting ing of cancer prognostic marker studies: association with reported
recurrence and progression by analysis of microsatellite loss of prognostic eVect. J Natl Cancer Inst 99:236243
heterozygosity in urine sediment and tumor tissue. Cancer Genet 52. Habuchi T, Marberger M, Droller MJ, Hemstreet GP III, Gross-
Cytogenet 167:1519 man HB, Schalken JA, Schmitz-Drger BJ, Murphy WM, Bono
35. Eissa S, Swellam M, Ali-Labib R, Mansour A, El-Malt O, Tash AV, Goebell P, Getzenberg RH, Hautmann SH, Messing E, Fradet
FM (2007) Detection of telomerase in urine by 3 methods: evalu- Y, Lokeshwar VB (2005) Prognostic markers for bladder cancer:
ation of diagnostic accuracy for bladder cancer. J Urol 178:1068 International Consensus Panel on bladder tumor markers. Urology
1072 66(6 Suppl 1):6474
36. Liu BC (2006) Urine telomerase and bladder cancer detection. 53. Lokeshwar VB, Habuchi T, Grossman HB, Murphy WM, Haut-
JAMA 295:998 mann SH, Hemstreet GP 3rd, Bono AV, Getzenberg RH, Goebell
37. Weikert S, Christoph F, Miller K (2006) Urine telomerase and P, Schmitz-Drger BJ, Schalken JA, Fradet Y, Marberger M,
bladder cancer detection. JAMA 295:998999 Messing E, Droller MJ (2005) Bladder tumor markers beyond
38. Gilbert SM, Veltri RW, Sawczuk A, Shabsigh A, Knowles DR, cytology: international consensus panel on bladder tumor markers.
Bright S, ODowd GJ, Olsson CA, Benson MC, Sawczuk IS Urology 66(6 Suppl 1):3563
(2003) Evaluation of DD23 as a marker for detection of recurrent 54. Altman DG, Lausen B, Sauerbrei W, Schumacher M (1994) Dan-
transitional cell carcinoma of the bladder in patients with a history gers of using optimal cutpoints in the evaluation of prognostic
of bladder cancer. Urology 61:539543 factors. J Natl Cancer Inst 86:829835
39. Bonner RB, Liebert M, Hurst RE, Grossman HB, Bane BL, Hem- 55. Schmoor C, Sauerbrei W, Schumacher M (2000) Sample size con-
street GP 3rd (1996) Characterization of the DD23 tumor-associ- siderations for the evaluation of prognostic factors in survival
ated antigen for bladder cancer detection and recurrence analysis. Stat Med 19:44152
monitoring. Marker Network for Bladder Cancer. Cancer Epidem- 56. Pajak TF, Clark GM, Sargent DJ et al (2000) Statistical issues in
iol Biomarkers Prev 5:971978 tumor marker studies. Arch Pathol Lab Med 124:10111015
40. Sawczuk IS, Pickens CL, Vasa UR, Ralph DA, Norris KA, Miller 57. Altman DG, Royston P (2000) What do we mean by validating a
MC, Ng AY, Grossman HB, Veltri RW (2002) DD23 biomarker. prognostic model? Stat Med 19:453473
A prospective clinical assessment in routine urinary cytology 58. Begg CB, Cramer LD, Venkatraman ES, Rosai J (2000) Compar-
specimens from patients being monitored for TCC. Urol Oncol ing tumour staging and grading systems: a case study and a review
7:185190 of the issues, using thymoma as a model. Stat Med 19:19972014
41. Shariat SF, Karam JA, Lerner SP (2008) Molecular markers in 59. Simon R, Altman DG (1994) Statistical aspects of prognostic fac-
bladder cancer. Curr Opin Urol 18:18 tor studies in oncology. Br J Cancer 69:979985
42. Sanchez-Carbayo M, Socci ND, Richstone L, Corton M, Behrendt 60. McGuire WL (1991) Breast cancer prognostic factors: evaluation
N, Wulkfuhle J, Bochner B, Petricoin E, Cordon-Cardo C (2007) guidelines. J Natl Cancer Inst 83:154155
Genomic and proteomic proWles reveal the association of gelsolin 61. Stockler MR, Boyd NF, Tannock IF (1998) Guide to studies of diag-
to TP53 status and bladder cancer progression. Am J Pathol nostic tests, prognostic factors, and treatments. In: Tannockand H (ed)
171:16501658 The basic science of oncology, 3rd edn. McGraw-Hill, New York
43. Mhawech-Fauceglia P, Fischer G, Alvarez V Jr, Ahmed A, Herr- 62. Drew PJ, Ilstrup DM, Kerin MJ, Monson JR (1998) Prognostic
mann FR (2007) Predicting outcome in minimally invasive (T1a factors: guidelines for investigation design and state of the art ana-
and T1b) urothelial bladder carcinoma using a panel of biomark- lytical methods. Surg Oncol 7:7176

123
World J Urol (2008) 26:511 11

63. Grossman HB (2000) The Wrst and second international work- 75. Chiong E, Gaston KE, Grossman HB (2007) Urinary markers in
shops on diagnostic and prognostic markers in bladder cancer. screening patients with hematuria. World J Urol. 4 December 2007
Urol Oncol 5:183184 [Epub ahead of print]
64. Goebell PJ, Groshen S, Schmitz-Drger BJ, Sylvester R, Kogevin- 76. Schmitz-Drger BJ, Tirsar LA, Schmitz-Drger C, Drsam J,
as M, Malats N, Sauter G, Grossman HB, Waldman F, Cote RJ Mellan Z, Bismarck E, Ebert T (2007) Immunocytology in the
(2004) The International Bladder Cancer Bankproposal for a assessment of patients with asymptomatic hematuria. World J
new study concept. Urol Oncol 22:277284 Urol. 12 December 2007 [Epub ahead of print]
65. Lotan/Malats this issue WJU 77. Schmitz-Drger BJ, Beiche B, Tirsar LA, Schmitz-Drger C,
66. Amling CL (2000) Diagnosis and management of superWcial blad- Bismarck E, Ebert T (2007) Immunocytology in the assessment of
der cancer. Curr Probl Cancer 25:219278 patients with asymptomatic microhaematuria. Eur Urol 51:1582
67. van der Poel HG, Debruyne FM (2001) Can biological markers re- 1588
place cystoscopy? An update. Curr Opin Urol 11:503509 78. Goodman GE, Thornquist MD, Edelstein C, Omenn GS (2006) Bi-
68. Heney NM (1992) Natural history of superWcial bladder cancer. orepositories: lets not lose what we have so carefully gathered!
Prognostic features and long-term disease course. Urol Clin North Cancer Epidemiol Biomarkers Prev 15:599601
Am 19:429433 79. Holland NT, PXeger L, Berger E, Ho A, Bastaki M (2005) Molec-
69. Holmng S, Johansson SL (2006) Impact of diagnostic and treat- ular epidemiology biomarkerssample collection and processing
ment delay on survival in patients with renal pelvic and ureteral considerations. Toxicol Appl Pharmacol 206:261268
cancer. Scand J Urol Nephrol 40:479484 80. van Gils MP, Stenman UH, Schalken JA, Schrder FH, Luider
70. Araki M, Nieder AM, Manoharan M, Yang Y, Soloway MS TM, Lilja H, Bjartell A, Hamdy FC, Pettersson KS, BischoV R,
(2007) Lack of progress in early diagnosis of bladder cancer. Urol- Takalo H, Nilsson O, Mulders PF, Bangma CH (2005) Innovations
ogy 69:270274 in serum and urine markers in prostate cancer current European re-
71. Bergman J, Reznichek RC, Rajfer J (2008) Surveillance of patients search in the P-Mark project. Eur Urol 48:10311041
with bladder carcinoma using Xuorescent in-situ hybridization on 81. Baker SG (2000) Identifying combinations of cancer markers for
bladder washings. BJU Int 101:2629 further study as triggers for early intervention. Biometrics
72. Tomera KM (2004) NMP22 BladderChek test: point-of-care tech- 56:10821087
nology with life- and money-saving potential. Expert Rev Mol Di- 82. Pepe MS (2000) Reciever operating characteristics methodology.
agn 4:78394 J Am Stat Assoc 95:308311
73. Agarwal PK, Black PC, Kamat AM (2007) Considerations on the 83. Swets JA, Pickett RM (1982) Evaluation of diagnostic systems:
use of diagnostic markers in management of patients with bladder methods from signal detection theory. Academic Press, New York
cancer. World J Urol Dec 19 [Epub ahead of print] 84. Pepe MS, Etzioni R, Feng Z, Potter JD, Thompson ML, Thorn-
74. Sarosdy MF, Kahn PR, ZiVer MD, Love WR, Barkin J, Abara EO, quist M, Winget M, Yasui Y (2001) Phases of biomarker develop-
Jansz K, Bridge JA, Johansson SL, Persons DL, Gibson JS (2006) ment for early detection of cancer. JNCI 93:10541061
Use of a multitarget Xuorescence in situ hybridization assay to diag-
nose bladder cancer in patients with hematuria. J Urol 176:4447

123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.