Tumour node metastasis staging of bladder cancer: prognosis

versus pitfalls
Nadine T. Gaisa, Corinna Henkel and Ruth Knüchel
Institute of Pathology, Medical Faculty, RWTH Aachen
University, Aachen, Germany
Correspondence to Ruth Knüchel, Professor and
Chairman of Pathology, Institute of Pathology,
University Hospital RWTH Aachen, Pauwelsstraße 30,
52074 Aachen, Germany
Tel: +49 241 8089280;
e-mail: rknuechel-clarke@ukaachen.de
Current Opinion in Urology 2010, 20:398–403
Purpose of review
The WHO classification of urothelial cancer in 2004 has made changes based on the
insights of molecular genetics, indicating bladder cancer with entities that are
genetically stable versus those that are genetically instable. Seen as work in progress,
the need of further validation is obvious. Clinical studies based on solid histological
diagnosis are as necessary as the definition of more molecular features of bladder
cancer.
Recent findings
Solid histological diagnosis includes sufficient clinical information and adequate tissue
processing. This combined with molecular data will lead to a more clear-cut distinction
between benign and malignant and possibly to another change in terminology with
higher concordance to other epithelial tumours. Whereas the identification of FGFR3
mutations has led to a better distinction of at least two pathways of urothelial
carcinogenesis, additional multiparametric approaches may help improve the still
inadequate search for urine and blood markers indicative of bladder cancer and/or its
progression. Proteomic profiling, sets of epigenetic markers, and micro RNAs will be
given as examples.
Summary
Recent data mainly support the concept of the WHO 2004 classification of bladder
cancer. We are optimistic that an even more clear-cut distinction between benign
recurring, nonprogressing tumours and more aggressive tumours will enable us to focus
and limit chemotherapy.

Keywords
bladder cancer, genetic instability, histopathology, multiparametric diagnosis, TNM

Curr Opin Urol 20:398–403

ß 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins
0963-0643

Histopathology of flat lesions

Introduction
In contrast to other tissues, for example testicular or
kidney tissue, the urothelium is a simple type of lining
epithelium that should allow a straight forward classifi-
cation of tumours. However, multifocality and recur-
rences have complicated the issue of adequate diagnosis
and prognostic assessment. Obviously the pathologist
can only assess the tissue obtained from the urologist,
and the urologist is dependant on what he sees during
cystoscopy. Whereas highly technological molecular
tools increasingly help to characterize tumour tissues
in more detail, the quality of molecular diagnosis and
even more individualized therapy approaches are depen-
dent on a high quality of both, the urology procedures
and the pathology processing and diagnosis. Some of the
pitfalls in pathology are described below. Furthermore,
the need for interaction between urologists and patho-
logists as well as the need of prospective studies
are emphasized.
0963-0643 ß 2010 Wolters Kluwer Health | Lippincott Williams & Wilkins
The WHO 2004 classification describes only two types of
precancerous flat lesions: dysplasia (intraurothelial neo-
plasia low grade ¼ slightly atypical cells and changes in
histoarchitecture without significant inflammation) and
carcinoma in situ (CIS; intraurothelial neoplasia high gra-
de ¼ highly atypical cells with varying degree of changes in
tissue architecture). These lesions should be clearly sep-
arated from reactive changes due to inflammation and
therapy. Whereas flat urothelial lesions have been reclas-
sified in the current WHO classification, there is also a
change due to the increasing application of fluorescence
cystoscopy with 5-aminolevulinic acid and its derivatives,
which unequivocally improves the detection of flat lesions
[1]. A result of increased detection should lead to critical
reflection of previous data. Are recurrences after white
light endoscopy due to overlooked lesions, is progression
due to overlooked lesions? Consequently a re-evaluation of
the biological and clinical role of flat lesions of the urothe-
lium (CIS > dysplasia) seems mandatory and should be
DOI:10.1097/MOU.0b013e32833c9649

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 TNM staging of bladder cancer Gaisa et al. 399

Figure 1 Important diagnostic aspects and pitfalls of histological diagnosis in bladder cancer. Histological sections stained
haematoxylin–eosin

(a) Flat intraurothelial lesion high grade ¼ carcinoma in situ of the urothelium. The arrow points to polymorphic nuclei in the urothelium surrounded by
regenerative flat urothelium. Abundant capillaries and lymphocytes are seen in the adjacent stroma. (b) Pseudoinvasion of a noninvasive papillary
urothelial carcinoma, low grade. The tangential section leads to the aspect of plump invasion (arrow) of the urothelial cells into the stroma. Serial
sections help to prove the lack of invasion. (c) True stromal invasion of an invasive papillary urothelial carcinoma, high grade. The irregular shape of the
tumour nests and the small clusters of cells help to make the decision of early stromal invasion. The diagnosis of early stromal invasion should also be
supported by serial sections of the tissue block. (d) Lymphangiosis carcinomatosa ¼ tumour cells in lymph vessels (arrow) in a case of small cell cancer
of the urothelium, an aggressive variant of urothelial bladder cancer.

carried out in prospective studies or similar retrospective
analyses as described by Tilki et al. [2] using merely cases
initially monitored by fluorescence endoscopy. The task of
the pathologist is the high-quality diagnosis of flat urothe-
lial lesions high grade (CIS). Since the lesion is defined by
highly aberrant – genetically instable – cell(s), serial
sections of the quite often denuding (¼ atypical cells
are shed into the lumen/urine) lesions are more important
than molecular analysis (Fig. 1a). Parallel cytology will
increase the rate of detection, since cells which shed off the
urothelial basement membrane can still be detected in the
bladder wash specimen. Cytology samples or separately
done cytology reports should be submitted to the pathol-
ogist for optimal diagnostic efficacy. A careful sample
handling and processing will for sure reduce the variability
in diagnosis as calculated and presented by Lee et al. [3].
Histopathology of noninvasive papillary lesions and
neoplasms
The term ‘noninvasive papillary lesions and neoplasms’
includes a spectrum of pathological entities ranging from
benign lesions like papillomas (papillary structures lined
by normal looking urothelium) and inverted papillomas
(endophytic cords of urothelial cells invaginating into
the lamina propria), borderline lesions as the papillary
urothelial neoplasm of low malignant potential
(PUNLMP) to noninvasive papillary urothelial carci-
nomas [4]. Whereas papillomas are definitely considered
benign and have a defined genetic pattern [mere fibro-
blast growth factor receptor 3 (FGFR3) mutations] the
additional entity of PUNLMP was created in order to
classify a subgroup of exophytic papillary lesions.
PUNLMP shows more urothelial cell layers than papil-
loma and behaves rather benign, since it only recurs and
does not progress (for review see [4]). Histological
definition of the entity demands thorough embedding
and sectioning in order not to miss less differentiated
areas of the tumour.
The histological grade of noninvasive papillary urothelial
carcinomas displays significant prognostic information,
especially regarding tumour progression [5,6]. Actually,

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 400 Bladder cancer
they are therefore subdivided into a two-tier grading
system (low grade/high grade), which abolishes the for-
mer heterogeneous group of grade 2 tumours. This modi-
fication provides a better stratification of prognosis which
is also useful for subsequent treatment/follow-up. Low-
grade noninvasive papillary tumours show an orderly
appearance with slight variations in nuclear polarity, size
and shape and infrequent mitotic figures usually in the
lower half of the epithelium. They progress (about 10%)
or cause death (about 5%) less frequently [4,7]. High-
grade noninvasive papillary tumours appear disordered,
with altered cellular polarity, pleomorphic nuclei and
plenty of mitotic figures including atypical ones. These
changes in morphology reflect the genetic instability of
high-grade tumours, which also results in progression
rates up to 40% [4,7]. Since noninvasive papillary low-
grade and high-grade tumours of the urothelium are the
only ones in cancer diagnostics called carcinoma in spite
of absent invasion, we consider it necessary to adapt
the language to that of other types of cancer by naming
the lesions ‘intraurothelial neoplasia’ (low grade and high
grade) as suggested by van der Kwast [8] and done in
cervix, testis, intestine and many other tissues.
Given the great heterogeneity of noninvasive papillary
tumours in terms of genetic, morphological and clinical
behaviour, a reliable and more accurate histopathological
diagnosis is necessary for optimizing patients’ treatment
and outcome. Protocols for sample preparation of surgical
specimens should be modified accordingly. Transurethral
tumour resections (TURs) should be completely
embedded during histological processing. From each
tissue block routine diagnostic haematoxylin and eosin
sections should be cut on a minimum of three levels
(different depths). If invasiveness is suspected, the whole
block should be cut in levels. Invasive versus noninvasive
cancer is the most challenging pitfall in the diagnosis
of papillary lesions. In fragmented specimens of TURs,
nonperpendicular protrusions of the bottom parts of
the papillae are mimicking invasiveness (invasion ¼
neoplastic nests, clusters or single cells within the lamina
propria) and these tangential sections are hardly avoid-
able (Fig. 1b, c). Pathologists have to be aware of the
‘pseudoinvasion’ phenomenon and exclude a proper
invasion by multiple level sectioning and by additional
visualization of the basal membrane by periodic acid
Schiff (PAS) reaction.
No molecular marker of noninvasive bladder cancer
is currently validated for routine practice to improve
risk stratification. Nevertheless, several recent studies
suggested that mutations of the FGFR3 gene could
provide additional information regarding the prognosis.
Kompier and co-workers [9] showed that recurrences in
patients with mutated primary tumours were of lower
stage and grade compared to primarily wild type tumours.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Additionally, Burger et al. [10] could demonstrate
enhanced progression-free survival times for patients
with FGFR3-mutated tumours. Furthermore, a combi-
nation of cytology (only sensitive for high-grade lesions)
and FGFR3 mutation analysis in urine could represent an
efficient noninvasive surveillance strategy, as such an
approach would allow the detection of cytologically invis-
ible (i.e. low-grade lesions) or cystoscopically invisible
(i.e. lesions in ureter and renal pelvis) but mutant recur-
rences.
Histopathology of invasive urothelial and nonurothelial
tumours
The variable histology of infiltrating bladder cancer is
represented by naming 13 variants in the current WHO
2004 classification in addition to nonurothelial neoplasms
(squamous cell carcinoma, adenocarcinoma, small cell
carcinoma and nonepithelial tumours).
It already is, and will become increasingly important to
name and know these entities, since in parallel to
tumours of other organs subgroups have and may have
to be treated differently and patients might also profit
from biologicals as individualized tumour therapy, for
example for small cell cancer [11]. A number of subgroups
as small cell cancer, lymphoepithelial, sarcomatoid and
micropapillary variants are already known to have worse
prognosis than regular urothelial cancer and should be
subjected to neoadjuvant chemotherapy even in the rare
cases of early diagnosis (stage pT1). It is assumed that
certain variant differentiations (e.g. neuroendocrine
differentiation, small cell cancer) found in parts of urothe-
lial carcinomas determines the prognosis of the patient
and therefore the following therapy decisions. This
obliges the pathologist to do a thorough sampling of
TUR specimens even of large, obviously invasive
tumours.
Apart from the detection of variants, correct stage
assessment has to be carried out in pathology. This is
not only challenging for early lamina propria invasion
(see noninvasive papillary tumours), but also for muscle
invasion (pT2). In case of lamina propria invasion only,
the current opinion is to substage these tumours accord-
ing to their relationship to the muscularis mucosae
(above, at or below) as this seems to be of prognostic
significance [12]. But until now this substaging has not
become mandatory. Identification of the muscularis
mucosae and even the distinction between muscularis
mucosae and muscularis propria (detrusor muscle) in
TUR specimens can be difficult. Immunohistochemistry
for smooth muscle actin can be required in some cases.
Scattered muscle bundles in extensive infiltrative
tumours should be evaluated depending on their overall
distribution throughout the tumour on low power
magnification.
 TNM staging of bladder cancer Gaisa et al. 401

Sample processing of TUR specimens should be per-
formed as in noninvasive papillary tumours (see above).
In cystectomy specimens showing a tumour less than
2 cm in diameter the lesion should be completely
embedded. If the tumour exceeds 2–3 cm in diameter
a perpendicular cross-cutting and additional sampling of
areas with deepest invasion and relationship to adjacent
structures/organs should be performed. Furthermore,
sampling of several areas of tumour is important for
the detection of lymphovascular (L) and vascular (V)
invasion (Fig. 1d), which has a significant impact on
the clinical outcome [13,14]. The described quality of
sample handling will allow a reliable staging and establish
a valid source for further molecular studies.
Multiparametric tools to support or replace
histopathology for refined prognostic assessment
Tremendous efforts and costs have not led to an appro-
priate biomarker for early detection or indication of
progression for bladder cancer so far. Therefore, multi-
parametric tools are discussed at present to give a more
precise answer to the biomarker quest.
The emerging field of clinical proteomics comprises
several platform technologies that allow high throughput
and sensitive investigations. For example, mass spec-
trometry, 2D gel electrophoresis and protein arrays get
increasingly precise due to new instrumentation and
better understanding of the urgent need for standard
operation procedures in screening methods [15]. Differ-
ent 2D gel electrophoresis-based studies with technical
new [16] or old approaches [17] revealed a lot of candidate
proteins like BLCA, A-FABP or Bikunin which showed
interesting stage and grade-specific protein alterations in
urine and tissue [18,19]. Another tool to find interesting
marker panels is the so called screening approach using
surface-enhanced laser desorption/ionization (SELDI) or
matrix-assisted laser desorption/ionization (MALDI)
time of flight (TOF) technologies. Our group performed
a serum-based analysis using 105 bladder cancer patients’
sera and 96 control sera. We performed a cation exchange
prefractionation and analysed the results using multi-
dimensional data analysis tools. Bladder cancer could
be identified with an overall sensitivity of 94.1% and
specificity of 89.2% by comparing peak patterns. Low-
stage samples were definable from control samples (pTa:
95.5% sensitivity) and a differentiation of low-grade and
high-grade tumours was also possible (sensitivity of
77.3%) [20]. MALDI imaging combines histological
information with differential tumour-specific protein pat-
terns, based on MALDI TOF analysis of histological
slides (Fig. 2). As this technique expanded in the last
years, identification of differential expressed proteins out
of a histological slide is possible even on formalin-fixed

Figure 2 Matrix-assisted laser desorption/ionization imaging experiment of pTa high-grade versus low-grade bladder cancer
tumours

An automatic classification from pTa G1, 2 and 3 into HG and LG was performed. (a) Histological haematoxylin–eosin-stained slide with unclear
grading in marked position. (b) Sum spectra in the mass range from 2000 to 16 000 Da obtained from several bladder cancer patients with pTa HG
(dark grey) and pTa LG (light grey) tumours. Protein peaks that are marked with an asterisk were used in the support vector machine-based model to
distinguish between pTa HG and LG tumours. Additionally one peak was zoomed in. (c) After MALDI imaging of the uncertain area and classification of
obtained peaks, area was classified as HG automatically, classification results are color coded (dark grey ¼ HG; light grey ¼ LG). HG, high grade; LG,
low grade; MALDI, matrix-assisted laser desorption/ionization.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 402 Bladder cancer
paraffin-embedded (FFPE) tissue which is normally
excluded from proteomic research [21].
Alternatively, DNA is a favourable target for clinical
analysis due to its high stability and easy amplification
using polymerase chain reaction (PCR)-based tech-
niques. Epigenetic alterations of gene expression seem
to be one of the major explanations for the diversity of
cancer, which is not only defined by DNA mutations. It is
known that promoter methylation of tumour-suppressor
genes leads to silencing of these genes and tumour
progression occurs. After the extension of analysis
material to body fluids the use of chips to analyse epi-
genetic changes improved the multiparametric diagnostic
approach. Renard and co-workers [22] for example,
recently showed their specific workflow to the right
methylation marker panel in an understandable and
comprehensive way. They first detected epigenetically
altered target genes using a chip-based approach compar-
ing different cancerous and normal bladder cancer cell
lines and validated their methylation on FFPE tissue
from various stages and grades of bladder cancer and
nontumour tissue specimens. Further validation included
a total of 339 control urines and 157 urine samples from
bladder cancer patients. TWIST and NID2 gene meth-
ylation revealed sensitivity rates of 88 and 94% and
specificity rates of 94 and 91% in the urine training
and validation sets. Sensitivity for low-grade pTa
tumours reached 80–89% which is quite impressive in
comparison to urine cytology with sensitivities of less
than 24% for pTa low-grade tumours [23].
In addition to DNA methylation studies, micro RNAs
(miRNAs) gained researchers’ interest. MiRNAs com-
prise a post-transcriptional epigenetic mechanism that
regulates gene expression by silencing of mRNA trans-
lation which has a potential role in carcinogenesis [24].
Catto and colleagues [25] did an extensive miRNA
study comparing 78 normal and malignant tissue samples.
Interestingly miRNA upregulation occurred mainly in
high-grade noninvasive or invasive bladder cancer,
whereas low-grade papillary bladder cancer showed a
downregulation of 18 from 25 disregulated miRNAs.
Additionally, the authors were able to build up a pTa
low-grade and high-grade (together with invasive bladder
cancer)-specific miRNA profiling pattern and were able
to separate the different miRNA profiles with high
sensitivity (90–100%) and specificity (80–100%). In
the high-grade miRNA profile, miRNA 21 was among
the most prominent ones. It is known to negatively
regulate the p53 pathway and therefore a loss of cellular
control results [26]. In the low-grade miRNA group seven
of the downregulated miRNAs were predicted to bind
FGFR3 and it was postulated that miRNA downregula-
tion in this case might promote FGFR3 expression before
mutation occurs. These three examples of multipara-
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
metric approaches provoke optimism towards a relevant
‘pattern recognition’, that is more adequate to assess the
biological variety of tumours.
Conclusion
Tumour node metastasis (TNM) staging of bladder can-
cer is the gold standard for prognostic determination and
therapy concepts. The interaction of urologists and path-
ologists combined with a thorough sample processing to
avoid pitfalls are key features of high-quality diagnosis in
pathology. Mainly on the basis of high-quality diagnosis,
molecular techniques can advance patients’ prognostic
assessment as a supplement or replacement of histo-
patholgy. The multiparametric approach in light of the
power of computational analysis renders the best chances
for additional prognostic groups.
Acknowledgements
The authors thank Dr K. Lindemann-Docter for providing part of the
histological photos for Fig. 1.
References and recommended reading
Papers of particular interest, published within the annual period of review, have
been highlighted as:
 of special interest
 of outstanding interest
Additional references related to this topic can also be found in the Current
World Literature section in this issue (p. 449).
1 Witjes JA, Redorta JP, Jacqumin D, et al. Hexaminolevulinate-guided
fluorescence cystoscopy in the diagnosis and follow-up of patients with
nonmuscle-invasive bladder cancer. Review of the evidence and recommen-
dations. J Urol 2010 [Epub ahead of print].
2 Tilki D, Reich O, Svatek RS, et al. Characteristics and outcomes of patients
with clinical carcinoma in situ only treated with radical cystectomy: an
international study of 243 patients. J Urol 2010 [Epub ahead of print].
3 Lee Mc, Levin HS, Jones JS. The role of pathology review of transurethral
bladder tumour resection specimens in the modern era. J Urol 2010 [Epub
ahead of print].
4 Miyamoto H, Miller JS, Fajardo DA, et al. Noninvasive papillary urothelial
 neoplasms: the 2004 WHO/ISUP classification system. Pathol Int 2010;
60:1–8.
Recent review of histological and clinical information concerning noninvasive
papillary bladder tumours.
5 Sylvester RJ, van der Meijden APM, Oosterlinck W, et al. Predicting recur-
rence and progression in individual patients with stage Ta T1 bladder cancer
using EORTC risk tables: a combined analysis of 2596 patients from seven
EORTC trials. Eur Urol 2006; 49:466–477.
6 Fernandez-Gomez J, Solsona E, Unda M, et al. Prognostic factors in patients
with nonmuscle-invasive bladder cancer treated with bacillus Calmette-Guér-
in: multivariate analysis of data from four randomized CUETO trials. Eur Urol
2008; 53:992–1002.
7 Montironi R, Lopez-Beltran A. WHO classification of bladder tumours: a
summary and commentary. Int J Surg Pathol 2005; 13:143–153.
8 Kwast VanDer T. How to combine the molecular profile with the clinicopatho-
logical profile of urothelial neoplastic lesions. Scand J Urol Nephrol 2008; 42
(Suppl 218):15–184.
9 Kompier LC, van der Aa MNM, Lurkin I, et al. The development of multiple
bladder tumour recurrences in relation to the FGFR3 mutation status of the
primary tumour. J Pathol 2009; 218:102–112.
10 Burger M, van der Aa MNM, van Oers JMM, et al. Prediction of progression of
nonmuscle-invasive bladder cancer by WHO 1973 and 2004 grading and by
FGFR3 mutation status: a prospective study. Eur Urol 2008; 54:835–843.
11 Black PC, Brown GA, Dinney CPN. The impact of variant histology on the
outcome of bladder cancer treated with curative intent. Urol Oncol Semin
Original Investig 2009; 27:3–7.

12 Orsola A, Trias I, Raventos CX, et al. Initial high-grade T1 urothelial cell
carcinoma: feasibility and prognostic significance of lamina propria invasion
microstaging (T1a/b/c) in BCG-treated and BCG-nontreated patients. Eur
Urol 2005; 48:231–238.
13 Bolenz C, Herrmann E, Bastian PJ, et al. Lymphovascular invasion is an
independent predictor of oncological outcomes in patients with lymph node-
negative urothelial bladder cancer treated by radical cystectomy: a multi-
centre validation trial. BJU Int 2010 [Epub ahead of print].
14 Shariat SF, Svatek RS, Tilki D, et al. International validation of the prognostic
value of lymphovascular invasion in patients treated with radical cystectomy.
BJU Int 2010 [Epub ahead of print].
15 Sigdel TK, Sarwal MM. The proteogenomic path towards biomarker discov-
ery. Pediatr Transplant 2008; 12:737–747.
16 Orenes-Pinero E, Corton M, Gonzalez-Peramato P, et al. Searching urinary
tumour markers for bladder cancer using a two-dimensional differential gel
electrophoresis (2D-DIGE) approach. J Proteome Res 2007; 6:4440–4448.
17 Rotterud R, Malmstrom PU, Wahlqvist R, et al. The star chart to Ta bladder
cancer: an unsophisticated analysis of two-dimensional gel electrophoresis
proteome maps. Scand J Urol Nephrol 2010; 44:76–83.
18 Moreira JM, Ohlsson G, Gromov P, et al. Bladder cancer-associated protein, a
 potential prognostic biomarker in human bladder cancer. Mol Cell Proteomics
2010; 9:161–177.
Interesting biomarker candidates derived out of and extensive two-dimensional gel
electrophoresis approach.
19 Ohlsson G, Moreira JM, Gromov P, et al. Loss of expression of the adipocyte-
type fatty acid-binding protein (A-FABP) is associated with progression of
human urothelial carcinomas. Mol Cell Proteomics 2005; 4:570–581.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
TNM staging of bladder cancer Gaisa et al. 403
20 Schwamborn K, Krieg RC, Grosse J, et al. Serum proteomic profiling in
patients with bladder cancer. Eur Urol 2009; 56:989–997.
21 Aerni HR, Cornett DS, Caprioli RM. High-throughput profiling of formalin-fixed
paraffin-embedded tissue using parallel electrophoresis and matrix-assisted
laser desorption ionization mass spectrometry. Anal Chem 2009; 81:7490–
7495.
22 Renard I, Joniau S, van Cleynenbreugel B, 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 2009 [Epub ahead of
print].
Interesting article about biomarker candidates by means of DNA methylation –
from cell culture experiments to urine – with a clear laid out workflow.
23 Tsui KH, Chen SM, Wang TM, et al. Comparisons of voided urine cytology,
nuclear matrix protein-22 and bladder tumour associated antigen tests for
bladder cancer of geriatric male patients in Taiwan, China. Asian J Androl
2007; 9:711–715.
24 He L, Thomson JM, Hemann MT, et al. A microRNA polycistron as a potential
human oncogene. Nature 2005; 435:828–833.
25 Catto JW, Miah S, Owen HC, et al. Distinct microRNA alterations char-
 acterize high- and low-grade bladder cancer. Cancer Res 2009; 69:8472–
8481.
Important contribution in the field of miRNA biomarker discovery.
26 Papagiannakopoulos T, Shapiro A, Kosik KS. MicroRNA-21 targets a network
of key tumour-suppressive pathways in glioblastoma cells. Cancer Res 2008;
68:8164–819072.