Skip to main content

Full text of "Primary bladder adenocarcinoma versus metastatic colorectal adenocarcinoma: a persisting diagnostic challenge."

See other formats


Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



DIAGNOSTIC PATHOLOGY 



RESEARCH Open Access 



Primary bladder adenocarcinoma versus 
metastatic colorectal adenocarcinoma: 
a persisting diagnostic challenge 

Somak Roy 1 , Matthew A Smith 1 , Kathy M Cieply 1,2 , Marie B Acquafondata 1 and Anil V Parwani 1,3 



Abstract 

Aim: This study attempted to distinguish primary bladder adenocarcinoma (PBA) from metastatic colonic 
adenocarcinomas (MCA), which is a difficult diagnostic and clinical problem. 

Methods: Twenty-four cases of bladder adenocarcinomas (12 primary & 12 metastatic colorectal) were included in 
the study with urothelial carcinoma (UC) and colonic adenocarcinoma (CA) as controls. A panel of 
immunohistochemical (IHC) stains along with fluorescence in-situ hybridization (FISH), using the UroVysion probe 
set, was performed. 

Results: The majority of the PBAs presented with advanced disease. Enteric histologic subtype was the most 
common morphological variant. Strong nuclear with cytoplasmic-membranous staining of (3-catenin was seen in 
75% of MCA and only 16.7% PBA (<10% staining cells). Although abnormal nuclear staining with E-cadherin was 
seen in both PBA and MCA, it was more frequent in former. CK-7, CK-20, villin and CDX-2 stains were not helpful in 
distinguishing the two entities. FISH did not reveal any unique differences in chromosomal abnormality between 
the two groups. 

Conclusion: Although there was a statistically significant difference in (3-catenin and E-cadherin staining between 
two groups, we did not find any IHC or FISH marker that was specific for PBA. Distinction between PBA and MCA 
remains a diagnostic problem and clinical correlation is vital before rendering a diagnosis. 

Virtual slides: The virtual slides for this article can be found here: http://www.diagnosticpathology.diagnomx.eu/vs/ 
1393156268152357 

Keywords: Bladder, Adenocarcinoma, Primary, Metastatic, Colorectal, Beta-catenin, E-cadherin 



Introduction 

Primary bladder adenocarcinoma (PBA), histologically 
comprised entirely or almost entirely of malignant glan- 
dular elements, is a rare tumor accounting for 0.5-2% of 
all malignant vesical tumors [1-3]. This glandular tumor, 
like other variants, arises through a process of divergent 
differentiation in urothelial carcinoma, which is extensive 
enough to predominate as the only histological compo- 
nent [1,4-6]. It is more frequent in males in their sixth 
decade of life, presenting with hematuria and symptoms 



* Correspondence: parwaniav@upmc.edu 

department of Pathology, University of Pittsburgh Medical Center, 
Pittsburgh, PA 15232, USA 

3 Division of Pathology Informatics, Department of Pathology, 5230 Centre 

Avenue, Room WG 07, Pittsburgh, PA 15232, USA 

Full list of author information is available at the end of the article 

(3 BioMed Central 



attributable to bladder irritation [1,7]. Two-thirds of PBA 
arise in the bladder cavity, especially in the posterior 
wall and trigone, and Approximately one-third originate 
from urachal remnants near the dome and anterior wall 
of the bladder [7,8]. Based on this, it is broadly classified 
as non-urachal and urachal adenocarcinoma, respect- 
ively. Although both subtypes of PBA have remarkably 
similar histological and immunohistochemical features, 
urachal subtype requires specific diagnostic criteria, put 
forth by Sheldon et al. [9] and Mostofi et al [10]. PBAs 
usually have associated surface glandular metaplasia or 
cystitis glandularis in the surrounding urothelial lining; 
however this is often difficult to document on small blad- 
der biopsies and trans-urethral resection specimens due 
to ulcerated and cauterized epithelium. Approximately 



© 2012 Roy et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative 
Commons Attribution License (http://creativecommons.Org/licenses/by/2.0), which permits unrestricted use, distribution, and 
reproduction in any medium, provided the original work is properly cited. 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 2 of 9 



90% of the vesical tumors arising in extrophied bladders 
are adenocarcinomas [7]. Additional risk factors include 
Schistsoma infection, villous adenoma and cystocele. 
[2,11-13] PBA is an aggressive malignancy with a ten- 
dency to present at a higher stage and is associated with 
a worse overall survival [1,2,13]. 

The more common secondary bladder adenocarcinomas 
include hematogenous, lymphatic or direct spread of 
adenocarcinoma from surrounding organs, especially 
colo-rectum, female genital tract, prostate, and urothelial 
carcinoma with focal glandular differentiation. Metastatic 
colonic adenocarcinoma (MCA) accounts for approxi- 
mately one third of secondary bladder tumors and is 
virtually indistinguishable from PBA based on histomor- 
phology and ultrastructural features. Evidence of cystitis 
glandularis or intestinal metaplasia can be a helpful clue in 
establishing the diagnosis of primary bladder adenocarcin- 
oma; however, MCAs have been reported to colonize the 
surface urothelium and mimic in-situ glandular lesions 
[3,7,14,15]. 

There are relatively limited numbers of studies that 
have looked into the role of immunohistochemistry in 
differentiating PBA from secondary bladder carcinoma, 
specifically MCA [2,3,14-17]. The role of (3-catenin in 
differentiating PBA from MCA was first reported by 
Wang et al. [3] in 2001. They attempted to analyze ab- 
normal nuclear localization of p-catenin in colonic 
adenocarcinomas (CA) due to dysregulation of the wnt 
pathway, in the setting of PBA. Currently, there is lim- 
ited published data on the use of (3-catenin in differenti- 
ating PBA from MCA [2,3,14,18]. In this study, we 
analyzed the role of a panel of six IHC stains in differen- 
tiating PBA from MCA. We also attempted to detect 
any difference in underlying chromosomal abnormality 
by FISH using UroVysion probe set. 

Materials and methods 

The Institutional Review Board of University of 
Pittsburgh approved this study. Twenty- four cases of 
bladder adenocarcinomas diagnosed and treated at Uni- 
versity of Pittsburgh Medical Center between 1999 and 
2010 were identified which included 12 cases each of 
PBA and MCA. We also included 5 cases each of 
urothelial carcinoma (UC) and colonic adenocarcinoma 

Table 1 Details of immunohistochemical panel 

Antibody Type 
Cytokeratin 7 Rabbit-Monoclonal 



(CA) as controls in this study. Formalin-fixed, paraffin 
embedded tissue blocks along with hematoxylin and 
eosin (H&E) stained slides were retrieved from the 
archives. The pathologists, in order to confirm the diag- 
nosis and exclude urothelial carcinoma with focal glan- 
dular differentiation, reviewed the slides independently. 
Cystoscopic, colonoscopic and radiological findings were 
reviewed to confirm the origin, location and stage of the 
tumor. 

A panel of immunohistochemical stains was performed 
on formalin-fixed, paraffin embedded, unstained sections 
from each case (Table 1). Briefly, sections were cut at 
5\i and mounted on Superfrost Plus slides and dried. 
Sections were deparaffinized and hydrated to deionized 
water. Heat induced epitope retrieval using Citrate buffer 
(pH 6, DAKO, Carpinteria, CA) was performed followed 
by endogenous Peroxidase quenching using 3% Hydrogen 
peroxide (Fisher Scientific, Houston, TX) and blocking 
(CAS, Invitrogen, Carlsbad, CA) for 10 minutes. Slides 
were incubated with primary antibody for 30-45 minutes, 
secondary antibody (Mach 2 Mouse HRP, Biocare, 
Concord, CA) for 30 minutes and substrate chromogen 
(DAKO, Carpinteria, CA) for 5 minutes. Slides were 
washed with TBS buffer between the incubations and 
finally counterstained with Harris hematoxylin, dehy- 
drated, cleared and cover slipped. 

Cytoplasmic staining with membrane accentuation 
(CM) was considered the normal pattern of staining with 
(3-catenin. Nuclear (N) staining was considered abnormal. 
CM staining was the expected pattern for E-cadherin 
whereas nuclear (N) staining was abnormal. Apical brush 
border staining pattern was considered positive for villin. 
Nuclear immunoexpression was the expected pattern 
with CDX2 antibody and cytoplasmic staining was con- 
sidered positive for CK7 and CK20 (Table 1). 

A total of nine cases (5 PBA & 4 MCA) were available 
for Fluorescence in-situ hybridization (FISH), which was 
performed using the UroVysion probe set (Abbott 
Molecular, Inc., Des Plaines, IL). Five cases each of UC 
and MCA were used as controls. Briefly, formalin-fixed 
paraffin-embedded sections, were mounted, and serially 
sectioned at 5\i intervals. The area of interest was demar- 
cated using an H&E stained slide. FISH slides were 
deparaffinized in xylene, dehydrated with 100% ethanol 



Dilution Source 

1 :1 00 Biocare Medical, Concord, CA 



Cytokeratin 20 Mouse Monoclonal 1:200 Biocare Medical, Concord, CA 

CDX-2 Mouse Monoclonal 1:100 Biocare Medical, Concord, CA 

Villin Mouse Monoclonal 1:100 Biocare Medical, Concord, CA 

(3-catenin Mouse Monoclonal 1:200 DAKO, Carpinteria, CA 



E-cadherin 



Mouse Monoclonal 



1 :2000 



BD Transduction Labs, San Jose, CA 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 3 of 9 



Table 2 Clinical features of primary bladder adenocarcinoma 



Case No 


Age 


Sex 


Location 


Size 

(cm) 


Procedure 


Stage ¥ 


Follow-up 
(months) 


Outcome 


Residual disease 


1 


63 


F 


Posterior 


3.6 


Cy sto u reth recto my 


l4aN2M0 


33 


AWD £ 


Lymph node 
metastasis 


2 


54 


F 


Posterior & 
Lateral 


4 


Radical cystectomy 


T4bN2M0 


16 


NED 9 




3 


87 


M 


Posterior 


6 


Partial cystectomy 


T1N0M0 


13 


NED 




4 


53 


M 


Anterior 


3.6 


Partial cystectomy 


T3N0M0 


4 


NED 




5 


70 


M 


Posterior 


6.8 


TURBT* 


1 4N0IV10 


2 


DOD € 




6 


62 


M 


Posterior 


3.5 


TURBT* 


1 2bN0M1 


18 


AWD 


Lung and pleural 
metastasis 


7 


59 


M 


Anterior 


2.9 


Partial cystectomy 


T3aN0M0 


132 


NED 




8 


59 


M 


Anterior 


2.9 


Partial cystectomy 


T3aN0M0 


128 


NED 




9 


62 


M 


Anterior 


4 


Radical cystectomy 


T4N0M0 


36 


AWD 


Lung and pleural 
metastasis 


10 


80 


M 


Posterior & 
Anterior 


8 


Radical cystectomy 


T3aN1 MO 


2 


AWD 


Multiple bony 
metastasis 


11 


56 


M 


Posterior 


5 


Radical cystoprostatectomy 


T3aN2M0 


4 


NED 




12 


70 


M 


Multifocal 1 


0.3-0.7 


Radical cystoprostatectomy 


T2aN0M0 


2 


NED 





t - Tumor located predominantly in posterior wall with other tumor nodules in anterior and lateral walls, t - TURBT - Transurethral resection of bladder tumor. 
3 - NED - No evidence of disease. € - DOD - Died of disease. £ - AWD - Alive with disease. ¥ -Pathologic stage. 



and then pretreated for 30 min in 0.2N HCL Slides were 
then digested in protease solution at 37°C. The target 
slide and probe were co-denatured at 90 C for 12 min- 
utes and incubated overnight at 37°C in a humidified 
chamber. Post-hybridization washes were performed 
using 2XSSC/0.3% Igepal at 72°C for 2 minutes. Slides 
were air-dried in the dark and counterstained with DAPI I 
(Abbott Molecular, Des Plaines, IL). Analysis was per- 
formed using an Applied Imaging Workstation equipped 
with Chroma Technology filters containing band excitors 
for SpectrumOrange, FITC, DAPI. Only individual and 
well-delineated cells were scored. Overlapping cells were 
excluded from the analysis. Approximately 60 cells were 



analyzed in the targeted region. Statistical analysis was 
performed using SPSS vl9.0 (IBM, Armonk, NY). 

Results 

Clinicopathologic features of primary bladder 
adenocarcinoma 

The clinicopathologic characteristics of 12 cases of PBA 
are summarized in Tables 2 and 3. The age at presentation 
ranged from 53-87 years (mean age 64.6 years) with a 
male predominance [4,18]. The cases did not demonstrate 
any associated predisposing conditions (Schistosomiasis, 
cystocele, bladder exstrophy). The majority of the cases 



Table 3 Pathologic features of primary bladder adenocarcinoma 



Case No 


Histological diagnosis 


Mucin 


SRC 


CC/CG 


Villous 


1 


MD AdenoCa (enteric) 


+ 




+ 




2 


PD AdenoCa (enteric) 




+ 


+ 


+ 


3 


MD AdenoCa (enteric) 


+ 




+ 


+ 


4 


MD to PD AdenoCa (enteric) 


+ 


+ 






5 


MD AdenoCa (enteric) 










6 


PD AdenoCa (SRC) 


+ 


+ 


+ 


+ 


7 


MD AdenoCa (enteric & mucinous) 


+ 




+ 




8 


MD AdenoCa (enteric & mucinous) 


+ 




+ 




9 


MD AdenoCa (enteric) 






+ 




10 


PD AdenoCa (mucinous and SRC) 


+ 


+ 






11 


Mucinous AdenoCa (enteric) 


+ 


+ 


+ 




12 


MD AdenoCa (enteric) 






+ 





MD Moderately differentiated, PD Poorly differentiated, AdenoCa Adenocarcinoma, SRC Signet ring cell carcinoma, CC Cystitis cystica, CG Cystitis glandularis, Villous 
Villous adenoma. 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 4 of 9 



of PBAs (75%) presented at a high pathologic stage 
(> pT3). Four of 12 (33.3%) PBA cases had node positive 
disease and one patient (8.3%) presented with lung 
metastasis (Table 2). 

Follow-up data was available on all the patients 
(2 months to 11 years), which is summarized in 
Table 2. Seven patients (58.3%) were alive with no evi- 
dence of disease and four patients (33.3%) were alive 
with evidence of disease in the form of distant metastasis 
(metastasis to lung and bone). Of the patients with pro- 
gressive disease, case numbers 1, 9 and 10 developed 
metastatic disease at 33, 36 and 2 months from the time 
of diagnosis, respectively (Table 2). 

Ten of the 12 (83.3%) cases of PBA were of the enteric 
type with varying degrees of differentiation (Figure la). 
Approximately two-third of the PBA cases originated in 
the posterior wall of the bladder and l/3rd were seen to 



arise from the anterior wall. The average tumor size was 
4.3 cm. In case number 12, there were multiple tumor 
nodules (0.3-0.7 cm) and although the tumor nodules 
were predominantly in the posterior wall, additional 
nodules were present in the anterior and lateral bladder 
walls (multifocal). Histological variations included the 
presence of mucin (8/12, 66.7%, Figure lb) and signet 
ring cells (5/12, 41.7%, Figure lc). Cystitis cystica and 
cystitis glandularis were seen in 9/12 (75%, Figure Id, le) 
and villous adenoma was seen in 3/12 (25%) cases. All 
MCA cases were metastatic from the colon and rectum 
(Figure If) (Table 3). 

Immunohistochemical findings 

The results of immunohistochemical staining for the 
PBA cases are summarized in Tables 4 and 5. All PBA 
cases demonstrated CM staining with p-catenin stain 




Figure 1 a: Primary bladder adenocarcinoma, enteric type. Moderately differentiated malignant glands are seen with dirty luminal necrosis. 
Elongated, enlarged and hyperchromatic nuclei with prominent stratification line the malignant glands. (Hematoxylin & eosin, xlOO). b: Primary 
bladder adenocarcinoma, mucinous type. Scattered small groups of tumor cells with intracytoplasmic mucin are seen in a background of 
abundant mucinous material (Hematoxylin & eosin, x200). c: Primary bladder adenocarcinoma, signet ring cell type: Tumor comprised of diffuse 
sheets of signet ring cells infiltrating the bladder wall (Hematoxylin & eosin, x200). d: Cystitis cystica involving the surface urothelium with 
underlying invasive primary bladder adenocarcinoma (Hematoxylin & eosin, xlOO). e: Extensive cystitis glandularis, intestinal type seen adjacent to 
a focus of invasive primary bladder adenocarcinoma (not seen in this image) (Hematoxylin & eosin, x40). f: Metastatic colorectal adenocarcinoma. 
Moderately differentiated, infiltrating malignant glands with morphological features similar to primary bladder adenocarcinoma see in a 
(Hematoxylin & eosin, xlOO). 

V ) 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 5 of 9 



Table 4 Results of p-catenin and E-cadherin 
immunostaining of primary bladder adenocarcinoma, 
metastatic and primary colorectal adenocarcinoma and 
urothelial carcinoma 







PBA 


MCA 


CA 


uc 


Beta-Catenin 


CM+N 


2 (16.7%)* 


9 (75%) 


5 (100%) 


0 




CM only 


10 (83.3%) 


3 (25%) 


0 


5 (100%) 




Total 


12 


12 


5 


5 


E-cadherin 


CM+N 


7 (58.3%) 


3 (25%) 


0 


0 




CM only 


5 (41.7%) 


9 (75%) 


5 (100%) 


5 (100%) 




Total 


12 


12 


5 


5 



*-Nuclear staining seen in <10% cells. 

CA Colorectal adenocarcinoma, CM cytoplasmic staining with membranous 
accentuation, MCA metastatic colorectal adenocarcinoma, N nuclear, UC 
urothelial carcinoma. 



(Figure 2a). Two cases (16.7%) demonstrated focal 
nuclear staining in addition (<5% cells). Nine of 12 (75%) 
MCAs demonstrated strong nuclear and CM staining 
(Figure 2b) and remaining (25%) did not demonstrate 
nuclear staining. The control cases of CA showed strong 
and diffuse nuclear as well as CM expression of p-catenin 
in 4/5 (80%) cases and focal (20% cells) nuclear staining in 
one case. Nuclear staining was not seen in UCs (Table 4). 

CM and nuclear staining with E-cadherin was restricted 
to PBA (7/12, 58.3%) and MCA (3/12, 25%) only 
(Figure 2c, 2d). The difference in abnormal nuclear ex- 
pression of e-cadherin between PBA and MCA was not 
statistically significant (Fishers Exact test, p =0.2 14). How- 
ever, when PBA was compared to MCA and CA com- 
bined, the difference was statistically significant (Fishers 
Exact test, p=0.046). (Table 5) Apical brush border 
staining with villin was seen in all cases of PBA and 
11/12 (91.7%) cases of MCA (Figure 3a). CDX-2 expres- 
sion was documented in 10/12 (83.3%) of PBA and all 
cases of MCA (Figure 3b). Four (33.3%) cases of PBA 
and 1 (8.3%) case of MCA demonstrated strong CK7 
expression (Figure 3c). CK20 expression was seen in all 
cases of PBA and MCA (Figure 3d) (Table 5). 



Table 5 Results of Cytokeratin 7 & 20, villin and CDX-2 
immunostaining of primary bladder adenocarcinoma, 
metastatic and primary colorectal adenocarcinoma and 
urothelial carcinoma 



PBA MCA CA UC 



Cytokeratin 7 


4 (33.3%) 


1 (8.3%) 


0 


5 (100%) 


Cytokeratin 20 


12 (100%) 


12 (100%) 


5 (100%) 


2 (40%) 


Villin 


12 (100%) 


11 (91.7%) 


5 (100%) 


0 


CDX 2 


10 (83.3%) 


12 (100%) 


5 (100%) 


0 



CA Colorectal adenocarcinoma, CM cytoplasmic staining with membranous 
accentuation, MCA metastatic colorectal adenocarcinoma, UC urothelial 
carcinoma. 



Fluorescence in-situ hybridization results 

The most frequent chromosomal abnormalities detected 
by UroVysion FISH in PBAs were 9p21 homozygous 
loss (HL) (6/9, 67%) (Figure 4) followed by Polysomy 
(PL) pattern (3/9, 33%) (Figure 5). HL, PL and single 
chromosome gain of chromosome 3 (SG) were seen 
in 3/5 (60%), 1/5 (20%), and 1/5 (20%) of PBA cases, 
respectively. Secondary adenocarcinoma showed HL 
(3/4, 75%), PL (2/4, 50%) and SG (chromosome 7) (1/4, 
25%), respectively (Table 6). 

Discussion 

Primary bladder adenocarcinoma is a rare tumor com- 
prising no more than 2% of all primary vesical malig- 
nancies [1-3]. in our series, the male to female ratio 
was 5:1 with mean age of 64.6 years. Eight of 12 cases 
(66.7%) originated from the posterior wall. Four tumors 
were seen in the anterior wall near the dome. However, 
they did not meet the other required criteria to be classi- 
fied as urachal type adenocarcinoma. These findings were 
in accordance to published literature [2,3,7,8]. 

The average size of tumor documented in our study 
was 4.25 cm (range 0.7-6.8 cm). Histologically, PBA is 
usually well to moderately differentiated and most fre- 
quently are of the enteric type, comprised of glandular 
structures lined by cuboidal to columnar cells with 
basally located vesicular nuclei and prominent nucleoli. 
The cytoplasm is usually apical with or without mucin 
vacuoles, the former representing goblet cells found in 
benign and malignant colonic glands. Mitoses and dirty 
necrosis are frequently seen [1,3,4,7]. Varying degree of 
extracellular mucin may be seen with these tumors. The 
mucinous variant is comprised of abundant extracellular 
pools of mucin with floating tumor cells. Additionally, 
some of the tumors may also harbor signet ring cells 
with intracellular mucin, the latter often referred to as 
the signet ring cell variant of PBA [2,18]. 

In the current study, there was a predominance of 
moderately differentiated PBA (75%), all of which were 
of the enteric type. Extracellular mucin and signet ring 
cells were seen in 66.7% and 41.7% cases, respectively. 
Case #6 had predominance of signet ring cells (>75%) 
and was classified as signet ring cell adenocarcinoma. 
Cystitis cystica/glandularis was documented in three- 
quarters of the cases and a villous adenoma like lesion 
was seen in 25% cases. Other variants, such as hepatoid 
and clear cell types, were not included in this study. 

Secondary bladder adenocarcinoma, which is more fre- 
quent than PBA, is represented by either direct extension 
of tumor from surrounding organs like the rectum, colon 
and female genital tract or lymphatic and hematogenous 
spread of tumor. MCA is by far the most frequent type in 
SBA. It is important to distinguish PBA from MCA for 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 6 of 9 




Figure 2 a: Primary bladder adenocarcinoma. Strong cytoplasmic membranous staining with (3-catenin (DAB chromogen, xlOO). b: Metastatic 

colorectal adenocarcinoma. Strong nuclear staining in addition to cytoplasmic staining with (3-catenin (DAB chromogen, xlOO). c: Primary bladder 

adenocarcinoma. Strong nuclear staining with e-cadherin in addition to weaker cytoplasmic staining pattern. This pattern was more frequent in 

this group of tumor in contrast to metastatic colorectal adenocarcinoma (DAB chromogen x200). d: Metastatic colorectal adenocarcinoma. 

Prominent cytoplasmic membranous staining with e-cadherin. Note the absence of nuclear staining (DAB chromogen, x200). 
k ) 



disease staging (local versus metastatic), and manage- 
ment [2,3,16]. 

Abnormal localization of p-catenin has been well 
documented in colonic adenocarcinomas with mutation 



in APC tumor suppressor gene [19]. The wild-type 
APC functions to degrade free cytoplasmic (3-catenin, 
an important molecule in cadherin mediated cell-to- cell 
adhesion system. In patients with mutated or absent 




Figure 3 a: Primary bladder adenocarcinoma demonstrating apical brush border staining with villin. Very similar staining pattern was also 
seen in metastatic colorectal adenocarcinoma. (DAB chromogen, xlOO). b: Primary bladder adenocarcinoma demonstrating strong diffuse nuclear 
staining with CDX-2. Metastatic colorectal adenocarcinoma demonstrated same staining pattern. (DAB chromogen, x100). c: Primary bladder 
adenocarcinoma shows strong cytoplasmic staining with CK7. Overall, it was infrequently seen in glandular bladder tumors, however slightly 
more frequently in primary than metastatic adenocarcinomas. (DAB chromogen, x100). d: Primary bladder adenocarcinoma shows diffuse strong 
cytoplasmic positivity for CK20. This was also seen in majority of metastatic colorectal adenocarcinoma. (DAB chromogen, x100). 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 7 of 9 




Figure 4 Primary bladder adenocarcinoma. FISH using UroVysion 
probe set demonstrates complete loss of yellow signal in some cells 
(homozygous loss 9p21) and single chromosome 3 gain in fewer 
cells (>2 red signals) (Red - CEP3, Green - CEP7, Aqua - CEP17 and 
Yellow - LSI 9p21,x600). 



APC, (3-catenin translocates to the nucleus of the cell 
and acts as a transactivating factor for the transcrip- 
tional factor Tcf-4 to regulate the expression of a num- 
ber of downstream target genes that are believed play a 
role in oncogenesis [20]. Wang et al. first exploited this 
finding in the setting of PBA and observed absence of 
nuclear staining in all PBA and nuclear staining in 81% 
MCA. This was reported to be a helpful distinction 
between the two entities [3]. Subsequently, few other 
studies reported the utility of p-catenin staining in 




Figure 5 Primary bladder adenocarcinoma. FISH using UroVysion 
probe set demonstrates a polysomy pattern with more than 2 red, 
green and aqua signals. (Red - CEP3, Green - CEP7, Aqua - CEP17 
and Yellow - LSI 9p21,x600). 

V / 



Table 6 Results of FISH using UroVysion probe set 


Case# 




FISH scores 




SG 


PL 


HL 


PBA 








Case 1 


0 


0 


4% 


Case 2 


0 


0 


12% 


Case 3 


8% (Ch 3) 


0 


4% 


Case 4 


0 


8% 


0 


Case 5 


0 


0 


0 


MCA 








Case 1 


0 


84% 


4% 


Case 2 


0 


52% 


4% 


Case 3 


0 


0 


8% 


Case 4 


4% (Ch 7) 


0 


0 



SG Single chromosome gain; PL Polysomy; HL Homozygous loss 9p21 
PBA Primary bladder adenocarcinoma; MCA Metastatic colonic 
adenocarcinoma Ch Chromosome; FISH fluorescence in-situ hybridization. 



settings of signet ring cell adenocarcinomas of bladder 
and urachal adenocarcinomas [2,14,18]. In our study, 
nuclear staining was predominantly seen in MCA and 
CA unlike PBA. In two cases of PBA nuclear staining 
was seen in less than 5% of the tumor cells. These find- 
ings are similar to the previously reported study by 
Gopalan et al. [14] who reported focal nuclear staining 
(15% cells) in 1 of 24 cases of urachal adenocarcinomas. 
However, the nuclear staining was seen in the clear cell 
urothelial component of the tumor rather than the 
adenocarcinoma component itself. Overall the differ- 
ence in p-catenin staining pattern between PBA and 
MCA was statistically significant (jfi test, p=0.001). 

The e-cadherin molecule forms an important compo- 
nent of the cell-cell adhesion complex in association 
with other proteins including catenin molecules (a-, p-, 
and y) and pl20 [21]. Loss of membranous localization 
of e-cadherin has been associated with invasion, metasta- 
sis and aggressive behavior in many human malignancies 
[21,22]. Loss of membranous staining with abnormal 
nuclear localization has been reported in pituitary, pan- 
creatic, esophageal and urothelial tumors [21-24]. The 
mechanism of this abnormal localization is still unclear. 
Loss of membranous e-cadherin expression in urothelial 
carcinoma has been correlated with aggressive disease, 
increased risk of nodal metastasis and death [22]. Tho- 
mas et al. reported loss of membranous e-cadherin ex- 
pression in approximately one-third of foci of PBA 
demonstrating signet-ring cell morphology in compari- 
son to colonic type PBA; however, they did not encoun- 
ter abnormal nuclear localization of e-cadherin [2]. 

In the current study, none of the CA or UC con- 
trols demonstrated nuclear staining pattern. E-cadherin 
molecule is comprised of 3 domains - extracellular, 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 8 of 9 



transmembrane and intracytoplasmic. Amongst the com- 
mercially available antibody clones, 36/E clone is directed 
against the cytoplasmic portion of e-cadherin. Our 
results of e-cadherin are based on use of this clone. 
(Table 1) The same clone was also used by the prior 
studies, which reported aberrant nuclear localization of 
e-cadherin [21-24]. 

CDX-2 is a mammalian homeobox gene, encoding a 
nuclear transcription factor, which is implicated as a 
tumor suppressor [16]. Nuclear staining is seen in nor- 
mal colonic epithelial cells and colonic adenocarcinoma 
has been reported [3,25]. CDX-2 expression was seen in 
83.3% cases of PBA, all cases of MCA and CA which is 
in accordance with the prior studies [3,14-16,25]. Inter- 
estingly, Suh et al. [16] also reported absence of CDX-2 
expression in half of the cases of PBA. Two cases 
(16.7%) of PBA in our series were negative for CDX-2. 

Villin is a 93-kilodalton actin-binding protein involved 
in the maintenance of brush border apparatus. Its ex- 
pression can be seen in gastrointestinal tract and renal 
epithelial cells as well as in adenocarcinomas of gastro- 
intestinal tract, pancreas, endometrium and ovary [16]. 
A limited number of studies have analyzed the role of 
villin in distinguishing PBA from SBA which reported 
overlapping staining between PBA and MCA [16-18]. 
We too did not find any difference in staining pattern 
between the two tumor groups. 

The CK7 and CK20 staining profile has been used pre- 
viously in distinguishing tumors from different sites. 
Gastrointestinal tumors, including colonic and rectal 
adenocarcinomas, tend to be CK20+ and CK7-. For 
urothelial neoplasms, the staining profile is the exact 
opposite. This difference was studied previously in PBA, 
but with limited success. Many PBA have been reported 
to show a CK20+ and CK7- profile, similar to colonic 
adenocarcinomas [3,15,17]. CK7 immunoexpression was 
seen in all cases of urothelial carcinoma and one third of 
PBAs. In contrast, only one (8.33%) case of SBA was 
positive for CK7 and all CAs were negative. CK20 immu- 
nostaining was seen in all PBAs, SBAs and colonic 
adenocarcinoma. Based on these findings, CK7 appeared 
to be of limited utility in distinguishing PBAs from sec- 
ondary bladder adenocarcinomas and CK20 was not able 
to distinguish between the two tumor groups. 

We also attempted to study the molecular aspect of 
PBA and MCA with a limited number of available cases 
(5 PBA and 4 MCA). FISH analysis using UroVysion 
probe sets were performed. To the best of our know- 
ledge, our study is the first one that utilized molecular 
methods to attempt to distinguish PBA from MCA. In 
our series of nine cases, the predominant abnormality 
was homozygous loss of 9p21 followed by polysomy pat- 
tern and single chromosome gain (chromosomes 3, 7). 
These findings however did not appear to be helpful in 



distinguishing PBA from SBA. Kipp et al [26]. studied 
FISH findings using UroVysion probe sets in bladder 
carcinoma variants. In their study, the most frequent 
abnormality documented in bladder adenocarcinoma 
was 9p21 homozygous loss followed by single gain only. 
They did not find significant difference in the chromo- 
somal abnormalities between the different tumor types. 
This study, unlike ours, did not compare bladder adeno- 
carcinoma with metastatic adenocarcinoma. 

Conclusion 

In summary, PBA is a rare and aggressive malignant 
neoplasm of the urinary bladder that morphologically 
mimics metastatic colorectal adenocarcinoma at the mor- 
phological level and to some extent immunohisto- 
chemically. A panel of antibodies comprised of (3-catenin, 
e-cadherin, CK7 and CDX-2 can be helpful in distin- 
guishing PBA from MCA in the light of appropriate 
clinical findings. Although FISH using the UroVysion 
probe set demonstrated various chromosomal abnormal- 
ities, it did not show significant differences between PBA 
and MCA. The results of the study reemphasize the per- 
sistence of the diagnostic challenge in distinguishing 
PBA and MCA clinico-pathologically. However, small 
cohort size was an important limitation of this study and 
therefore larger studies are required to substantiate the 
above conclusion. Moving further with question, It may 
be of interest to investigate the molecular landscape of 
the two tumors using high-throughput molecular ana- 
lysis methods for possible diagnostic biomarkers. 

Competing interests 

The authors declare that they have no competing interests. 
Authors' contribution 

MBA performed immunohistochemistry on all the cases in this study. KMC 
performed FISH studies on all these cases. MAS helped in case organization 
and critical manuscript review. SR wrote the manuscript, reviewed 
immunohistochemistry and FISH results and analyzed the data, AVP 
reviewed the immunohistochemistry and FISH results, critically reviewed the 
paper, arranged resources for the project and uploaded the manuscripts. All 
authors read and approved the final manuscript. 

Author details 

department of Pathology, University of Pittsburgh Medical Center, 
Pittsburgh, PA 15232, USA. 2 Florescence in-situ hybridization research and 
development laboratory, University of Pittsburgh Medical Center, Pittsburgh, 
PA 15232, USA. 3 Division of Pathology Informatics, Department of Pathology, 
5230 Centre Avenue, Room WG 07, Pittsburgh, PA 15232, USA. 

Received: 28 September 2012 Accepted: 22 October 2012 
Published: 2 November 2012 

References 

1. Grignon DJ, Ro JY, Ayala AG, Johnson DE, Ordonez NG: Primary 
adenocarcinoma of the urinary bladder. A clinicopathologic analysis of 
72 cases. Cancer 1991, 67:2165-2172. 

2. Thomas AA, Stephenson AJ, Campbell SC, Jones JS, Hansel DE: 
Clinicopathologic features and utility of immunohistochemical markers 
in signet-ring cell adenocarcinoma of the bladder. Hum Pathol 2009, 
40:108-116. 



Roy et al. Diagnostic Pathology 201 2, 7:1 51 
http://www.diagnosticpathology.org/content/771/151 



Page 9 of 9 



10. 



12. 



14. 



16. 



17. 



19. 



20. 



21. 



22. 



23. 



24. 



25. 



Wang HL, Lu DW, Yerian LM, Alsikafi N, Steinberg G, Hart J, Yang XJ: 
Immunohistochemical distinction between primary adenocarcinoma of 
the bladder and secondary colorectal adenocarcinoma. Am J Surg Pothol 
2001,25:1380-1387. 

Anderstrom C, Johansson SL, von Schultz L: Primary adenocarcinoma of 
the urinary bladder. A clinicopathologic and prognostic study. Cancer 
1983,52:1273-1280. 

Jones WA, Gibbons RP, Correa RJ Jr, Cummings KB, Mason JT: Primary 
adenocarcinoma of bladder. Urology 1 980, 15:1 19-122. 
Nocks BN, Heney NM, Daly JJ: Primary adenocarcinoma of urinary 
bladder. Urology 1 983, 21:26-29. 

Epstein Jl, Amin MB, Reuter VE: Glandular lesions. Philadelphia: Lippincott 
Williams & Wilkins; 2004. 

Melicow MM: Tumors of the urinary bladder: a clinico-pathological 
analysis of over 2500 specimens and biopsies. J Urol 1955, 74:498-521. 
Sheldon CA, dayman RV, Gonzalez R, Williams RD, Fraley EE: Malignant 
urachal lesions. J Urol 1984, 131:1-8. 

Mostofi FK, Thomson RV, Dean AL Jr: Mucous adenocarcinoma of the 
urinary bladder. Cancer 1955, 8:741-758. 

Abol-Enein H, el Mekresh MM, el Baz MA, Abol-Enein H, Ghoneim MA: 
Primary adenocarcinoma of the urinary bladder: a report of 185 cases. 

Br J Urol 1998, 82:206-212. 

Shaaban AA, Elbaz MA, Tribukait B: Primary nonurachal adenocarcinoma in 
the bilharzial urinary bladder: deoxyribonucleic acid flow cytometric and 
morphologic characterization in 93 cases. Urology 1998, 51:469-476. 
Sung MT, Lopez-Beltran A, Eble JN, MacLennan GT, Tan PH, Montironi R, 
Jones TD, Ulbright TM, Blair JE, Cheng L: Divergent pathway of intestinal 
metaplasia and cystitis glandularis of the urinary bladder. Mod Pathol 
2006, 19:1395-1401. 

Gopalan A, Sharp DS, Fine SW, Tickoo SK, Herr HW, Reuter VE, Olgac S: 
Urachal carcinoma: a clinicopathologic analysis of 24 cases with 
outcome correlation. Am J Surg Pathol 2009, 33:659-668. 
Raspollini MR, Nesi G, Baroni G, Girardi LR, Taddei GL: 
Immunohistochemistry in the differential diagnosis between primary 
and secondary intestinal adenocarcinoma of the urinary bladder. Appl 
Immunohistochem Mol Morphol 2005, 1 3:358-362. 
Suh N, Yang XJ, Tretiakova MS, Humphrey PA, Wang HL: Value of CDX2, 
villin, and alpha-methylacyl coenzyme A racemase immunostains in the 
distinction between primary adenocarcinoma of the bladder and 
secondary colorectal adenocarcinoma. Mod Pathol 2005, 18:1217-1222. 
Tamboli P, Mohsin SK, Hailemariam S, Amin MB: Colonic adenocarcinoma 
metastatic to the urinary tract versus primary tumors of the urinary tract 
with glandular differentiation: a report of 7 cases and investigation 
using a limited immunohistochemical panel. Arch Pathol Lab Med 2002, 
126:1057-1063. 

Del Sordo R, Bellezza G, Colella R, Mameli MG, Sidoni A, Cavaliere A: Primary 
signet-ring cell carcinoma of the urinary bladder: a clinicopathologic and 
immunohistochemical study of 5 cases. Appl Immunohistochem Mol 
Morphol 2009, 17:18-22. 

Chung DC: The genetic basis of colorectal cancer: insights into critical 
pathways of tumorigenesis. Gastroenterology 2000, 1 19:854-865. 
Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B, Kinzler KW: 
Activation of beta-catenin-Tcf signaling in colon cancer by mutations in 
beta-catenin or APC. Science 1997, 275:1787-1790. 
Elston MS, Gill AJ, Conaglen JV, Clarkson A, Cook RJ, Little NS, Robinson BG, 
Clifton-Bligh RJ, McDonald KL: Nuclear accumulation of e-cadherin 
correlates with loss of cytoplasmic membrane staining and invasion in 
pituitary adenomas. J Clin Endocrinol Metab 2009, 94:1436-1442. 
Byrne RR, Shariat SF, Brown R, Kattan MW, Morton RJ, Wheeler TM, Lerner 
SP: E-cadherin immunostaining of bladder transitional cell carcinoma, 
carcinoma in situ and lymph node metastases with long-term followup. 
J Urol 2001, 165:1473-1479. 

Chetty R, Serra S, Asa SL: Loss of membrane localization and aberrant 

nuclear E-cadherin expression correlates with invasion in pancreatic 

endocrine tumors. Am J Surg Pathol 2008, 32:41 3-41 9. 

Salahshor S, Naidoo R, Serra S, Shih W, Tsao MS, Chetty R, Woodgett JR: 

Frequent accumulation of nuclear E-cadherin and alterations in the Wnt 

signaling pathway in esophageal squamous cell carcinomas. Mod Pathol 

2008,21:271-281. 

Werling RW, Yaziji H, Bacchi CE, Gown AM: CDX2, a highly sensitive and 
specific marker of adenocarcinomas of intestinal origin: an 



immunohistochemical survey of 476 primary and metastatic carcinomas. 

Am J Surg Pathol 2003, 27:303-310. 
26. Kipp BR, Tyner HL, Campion MB, Voss JS, Karnes RJ, Sebo TJ, Hailing KC, 
Zhang J: Chromosomal alterations detected by fluorescence in situ 
hybridization in urothelial carcinoma and rarer histologic variants of 
bladder cancer. Am J Clin Pathol 2008, 130:552-559. 



doi:1 0.1 1 86/1 746-1 596-7-1 5 1 

Cite this article as: Roy et al:. Primary bladder adenocarcinoma versus 
metastatic colorectal adenocarcinoma: 

a persisting diagnostic challenge. Diagnostic Pathology 2012 7:151. 



Submit your next manuscript to BioMed Central 
and take full advantage of: 

• Convenient online submission 

• Thorough peer review 

• No space constraints or color figure charges 

• Immediate publication on acceptance 

• Inclusion in PubMed, CAS, Scopus and Google Scholar 

• Research which is freely available for redistribution 



Submit your manuscript at 
www.biomedcentral.com/submit 



o 



BioMed Central