fascin stain as a potential marker of invasiveness in carcinomas of the urinary bladder: a...

ORIGINAL ARTICLES

Fascin Stain as a Potential Markerof Invasiveness in Carcinomas ofthe Urinary Bladder:A Retrospective Study With Biopsyand Cytology CorrelationRyan McKnight, M.D., Cynthia Cohen, M.D.,and Momin T. Siddiqui, M.D., F.I.A.C.*

The evaluation of invasion in urothelial carcinomas of the urinarybladder cannot be determined on cytology and can be particularlychallenging in biopsy cases with limited sampling. Recent studiesof bladder resection specimens suggest that fascin overexpressionmay be a marker of aggressive urothelial carcinomas and canhelp facilitate the assessment of invasion. In this study, we eval-uated urine cytology and corresponding biopsy specimens withproven invasive urothelial carcinoma for fascin expression byimmunohistochemistry. Thirty-five patients diagnosed with posi-tive urine cytology and biopsy-proven invasive urothelial carci-noma between January 2003 and February 2009 were identified.We found increased fascin expression in 100% (35/35) of Sure-PathTM urine cytology preparations as well as 100% (35/35) ofcorresponding biopsy cases with invasive urothelial carcinoma.On urine cytology, cytoplasmic fascin staining was moderate tointense in malignant tumor cell clusters and single cells and notobserved in benign urothelial cells. Staining in biopsy cases wasgenerally intense and cytoplasmic and present in both the invasive(100%) and noninvasive (31%) components of the lesion. Thesefindings uphold the association of increased fascin expression ininvasive urothelial carcinomas of the urinary bladder. We further-more demonstrate that fascin staining can be performed success-fully on SurePathTM urine cytology preparations in whichincreased fascin expression correlates with invasion on biopsy.While not a definitive marker of invasion, as it is observed in insitu carcinoma, we conclude that the utilization of fascin immuno-histochemistry on urine cytology might serve as a useful adjunctin predicting invasiveness in subsequent biopsies. Diagn. Cytopa-thol. 2011;39:635–640. ' 2010 Wiley-Liss, Inc.

Key Words: fascin; urothelial carcinoma; bladder biopsy; urinecytology

Carcinomas of the urinary bladder account for *3.2% of

all cancers world wide with an estimated 336,000 new

cases each year.1 The prevalence of bladder tumors in

Western countries is significantly higher compared with

that in developing countries and is associated with signifi-

cant morbidity and mortality in the United States account-

ing for more than 14,000 deaths annually.1 Despite

increasing incidence, mortality rates have decreased over

the past 20 years with survival improvements owing

largely to early diagnosis and treatment.2

Although the initial pathological diagnosis of urothelial

carcinoma is made according to the WHO classification

from a biopsy obtained by cystoscopy or transurethral

resection of primary tumor, urine cytology plays an inte-

gral role in urothelial carcinoma screening as well as ini-

tial staging and risk assessment.3 Furthermore, urine cy-

tology is recommended in the assessment of response to

treatment and for long-term follow-up in patients once the

diagnosis has been established.3 Recent studies indicate

that the most powerful prognostic factor in patients with

urothelial carcinoma of the urinary bladder is tumor

stage.4,5 Subsequently one of the most important aspects

of tumor staging is the presence or absence of invasion of

tumor cells into the lamina propria, underlying muscularis

propria, and perivesicular tissue. Thus, the biologic

behavior and subsequent clinical outcome of such neo-

plasms are largely dependent on the presence and extent

of invasion into the bladder wall and surrounding tissues

which proves difficult to diagnose in some biopsy cases

and cannot be assessed on urine cytology.

Department of Pathology and Laboratory Medicine, Emory UniversityHospital, Atlanta, Georgia

*Correspondence to: Momin T. Siddiqui, M.D., F.I.A.C., Departmentof Pathology, Room G179B, Emory University Hospital, 1364 CliftonRoad NE, Atlanta, GA 30322. E-mail: [email protected]

Received 5 March 2010; Accepted 5 April 2010DOI 10.1002/dc.21429Published online 14 October 2010 in Wiley Online Library

(wileyonlinelibrary.com).

' 2010 WILEY-LISS, INC. Diagnostic Cytopathology, Vol 39, No 9 635

Fascin is a 55-kDa actin-bundling protein that plays an

important role in cell motility, migration, and adhesion6

Three forms of fascin exist in vertebrates, with recent

studies suggesting that fascin-1 protein level is signifi-

cantly increased in transformed epithelial cells and several

common carcinomas.7 To our knowledge, only two stud-

ies exist regarding fascin expression in urothelial carcino-

mas of the urinary bladder, the most recent of which sug-

gests an association between fascin overexpression and

urothelial carcinoma invasiveness.7,8 Neither of these

studies, however, has assessed fascin staining on urine cy-

tology, and this is the first study to utilize fascin immuno-

histochemistry in urine cytology with biopsy correlation.

Materials and Methods

Patients and Specimens

With Investigational Review Board (IRB) permission, the

surgical pathology and cytopathology files of Emory

University Hospital were searched for all positive urine

cytology cases over a six year period (January 2003 to

February 2009). We selected 35 patients with concurrent

biopsy proven invasive urothelial carcinoma. Fascin immu-

nohistochemical staining on SurePathTM urine cytology

samples and corresponding biopsy was performed. Fascin

staining was also performed on five negative control urine

cytology and biopsy cases. Selection criteria included cases

with quantitatively adequate material on both cytology and

biopsy cases. In biopsy cases, special care was taken to

select cases with definitive evidence of invasion, good

architectural preservation, and minimal cautery artifact.

Immunohistochemistry

For biopsy specimens, sections (5 lm) of formalin-fixed,

paraffin-embedded material were deparaffinized and rehy-

drated to deionized water. Antigen retrieval was performed

in citrate buffer (pH 6.0), using an electric pressure cooker

for 3 minutes at 12–15 pounds per square inch (PSI)

(*1208C), and cooled for 10 minutes prior to immunostain-

ing. All slides were loaded on an automated system (Dako

AutoStainer plus, Carpinteria, CA) and exposed to 3% H2O2

for 5 minutes, incubated with primary monoclonal fascin

antibody (Dako) (dilution 1:160) for 30 minutes, with

labeled polymer (Envision1 + dual link) for 30 minutes,

3,30-diaminobenzidine (DAB) as chromogen for 5 minutes,

and hematoxylin as counterstain for 5 minutes. These

incubations were performed at room temperature; between

incubations sections were washed with Tris-buffered saline

(TBS). Cover-slipping was performed using the Tissue Tek

SCA (Sakura Finetek USA, Inc.) coverslipper.

On initial processing SurePathTM urine cytology slides

(BD-TriPath Imaging, Burlington, NC) were prepared by

centrifuging for 10 minutes at 1,866 rpm. The supernatant

was decanted, and one full pipette of CytoRich Red

(Thermo Scientific, Waltham, MA) was added to the

resulting cell pellet, followed by allowing the specimen to

sit for at least 30 minutes. The resulting suspension was

centrifuged for 10 minutes at 1,866 rpm. The remaining

pellet was vortexed, and Tris-buffered water was added to

the 10-ml line. This suspension was then centrifuged for

5 minutes at 1,866 rpm to concentrate the specimen. The

remaining pellet was vortexed and placed in a correspond-

ing slot on the AutoCyte PREP system slides (BD-TriPath

Imaging, Burlington, NC). Coated glass slides with their

accompanying settling chambers were placed onto stain-

ing racks. The ‘‘non-gyn’’ program was run, and the

instructions prompted by the computer were followed.

Once staining was completed, the slides were removed

and dipped in reagent alcohol above the preparation area

of each slide. The slides were cleared in xylene and then

coverslipped. The slides were later decoverslipped by

immersion in xylene, and rehydrated in graded alcohols.

Immunohistochemical staining was performed beginning

with antigen retrieval as noted above. The slides were not

exposed to H2O2 for 5 minutes in the interest of maintain-

ing cellular integrity and preventing tissue damage.

Positive controls of anaplastic large cell lymphoma as

well as negative controls with primary antibody replaced

with TBS, were run with the patient/study slides.

Scoring

Biopsy cases were scored as positive for fascin if cytoplas-

mic staining was observed. Endothelial cells served as inter-

nal controls in biopsies with preserved vascular structures.

Intensity of immunostaining was categorized using endothe-

lial cells as a reference into weak (1+), moderate (2+), and

intense staining (3+) categories. The intense and weak stain-

ing categories included cases that exhibited greater and

lesser degrees of immunoreactivity compared with endothe-

lial cells respectively, while the moderate staining category

included cases that exhibited an equal degree of immunore-

activity to endothelial cells. Biopsy cases were further cate-

gorized into focal (less than 25%), intermediate (25–75%),

and diffuse (greater than 75%) categories based on the per-

centage of positively stained neoplastic cells.

Urine cytology cases were scored as positive for fascin

if cytoplasmic staining was observed in malignant cell

clusters and single cells.

Statistical Analysis

The predictive capacity of fascin staining on urine cytol-

ogy specimens was determined using sensitivity and spec-

ificity calculations with biopsy diagnosis of the presence

or absence invasion designated as the gold standard.

Results

All 35 biopsy cases were classified as high-grade invasive

urothelial carcinomas. Twenty-six cases (26/35) were pT1

carcinomas with lamina propria invasion while nine cases

MCKNIGHT ET AL.

636 Diagnostic Cytopathology, Vol 39, No 9

Diagnostic Cytopathology DOI 10.1002/dc

(9/35) were at least pT2 carcinomas with invasion of the

muscularis propria. Eight pT1 cases were devoid of muscula-

ris propria, precluding the evaluation of muscularis propria

invasion. While the majority of cases were papillary urothe-

lial carcinomas (Fig. 1A), other histologic variants included

two poorly differentiated carcinomas (2/35), one small cell

carcinoma (1/35), one papillary urothelial carcinoma with

squamous differentiation (1/35), and one papillary urothelial

carcinoma with micropapillary features (1/35).

Immunohistochemistry on SurePathTM CytologyPreparations

Intense cytoplasmic staining was observed in malignant

cell clusters in all thirty-five SurePathTM urine cytology

Fig. 1. A: Clusters of high-grade urothelial carcinoma including cells with high nuclear to cytoplasmic ratio and hyperchromatic nuclei (Papanicolaou,360). B: A group of malignant cells highlighted by fascin stain (arrow). Background benign urothelial cells (arrowhead) and squamous cells showing noreactivity (Papanicolaou, 340). C: Fragments of high-grade urothelial carcinoma (arrow) and single cells (arrowhead) demonstrating fascin reactivity(Papanicolaou, 360). D: Group of malignant cells and single cells demonstrating fascin reactivity. Background shows benign urothelial cells and cellulardebris (Papanicolaou, 360). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

UROTHELIAL CARCINOMA

Diagnostic Cytopathology, Vol 39, No 9 637


cases (sensitivity ¼ 100%). Intense staining was not

detected in benign urothelial cells in which only occa-

sional faint, nonspecific cytoplasmic and nuclear staining

was observed (Figs. 1B–1D). No staining was observed in

background inflammatory or squamous cells (Figs. 1B

and D). All five negative control urines were also nega-

tive for fascin staining (specificity ¼ 100%).

Immunohistochemistry on Biopsy Specimens

All 35 (100%) corresponding biopsy cases with invasive

urothelial carcinoma were positive for fascin (Figs. 2A–C).

Vascular endothelial cells were positive for fascin

(Fig. 2B), and fascin staining was weak to absent in areas

with cautery artifact. When positive in neoplastic urothe-

lium, fascin staining was generally intense and cytoplasmic

and present in both invasive and noninvasive neoplastic

areas with 31 (89%) cases demonstrating intense

immunostaining and four cases (11%) demonstrating mod-

erate staining. Additionally, fascin immunoreactivity was

generally diffuse with diffuse staining observed in nineteen

(54%) cases; eight cases each demonstrated intermediate

(23%) and focal staining (23%) respectively. Weak to

moderate staining was noted in the basal layer but was

absent in the superficial urothelium of five negative

control biopsy cases. In eleven cases (31%), fascin

immunoreactivity was observed in both the invasive and

noninvasive components of the lesion examined. Interest-

ingly, in six cases (17%) fascin highlighted isolated inva-

sive tumor cells and was negative in the overlying non-

invasive tumor component (Figs. 2A and C). Fascin also

highlighted tumor cell clusters in lymphovascular channels

in three cases with evidence of lymphovascular invasion

(Fig. 2B).

Discussion

Clinical outcomes of urinary cancer are associated with a

variety of biologic, molecular, and genetic features.9–11 A

number of new methodologies involving potential urinary

markers have been developed as non-invasive adjuncts to

conventional techniques for the diagnosis of urothelial

carcinoma including bladder tumor antigen, NMP221,

fibrin degradation products, telomerase, fluorescence in

situ hybridization test, and flow cytometry.9–11 Despite

these advances, conventional methods are still the main-

stay in making the diagnosis. Although the prevalence of

urothelial carcinoma makes cytologic screening in the

general population unfeasible, screening is generally rec-

ommended in patients with known risk factors or unex-

plained urologic symptoms.12

Although limited by a sensitivity of *50%, urine cy-

tology is highly specific and remains an important meth-

odology in the diagnosis and treatment of urothelial carci-

nomas of the urinary bladder.12,13 While cystoscopic bi-

opsy of visible lesions is generally considered more

sensitive than urine cytology in most cases, clinical limi-

tations also exist as evidenced by operator-dependent vari-

ability in visualization of lesions and biopsy technique.

The utility of cystoscopic biopsy is also limited in instan-

ces where there is no specific lesion to biopsy such as in

surveillance after tumor resection and if structural abnor-

malities such as diverticulae mask visualization. These

limitations provide the impetus for the utilization of urine

cytology in conjunction with bladder biopsy.

Prognostically, many investigators have shown a statis-

tically significant survival difference between patients

with invasive and non invasive tumors.4,5 Furthermore,

there are significant differences in treatment regimens

depending on the depth of invasion observed on a biopsy

specimen. Thus, the pathologist must be cautious in mak-

ing an invasive diagnosis on biopsy. Unfortunately, bi-

opsy interpretation is inherently susceptible to interob-

server variability and is complicated by the lack of stand-

ardized criteria for identifying lamina propria invasion.12

Technical limitations also exist including tangential orien-

tation, and crush and cautery artifact brought about by the

biopsy procedure itself.12 Given the high specificity of

urine cytology with regard to this diagnosis, the use of

ancillary methods to predict the propensity for invasion

might be useful in conjunction with bladder biopsy.

Fascin, an actin-bundling protein with various func-

tions, induces membrane protrusions, enhances epithelial

cell motility, and is highly expressed in transformed epi-

thelial cells as well as in specialized normal cells.14 A

number of studies have demonstrated high levels of fascin

expression in human malignancies, in particular carcino-

mas and lymphomas.15–18 Given the role fascin plays on

cell motility, investigations to characterize its effect on

carcinogenesis have suggested its potential role in tumor

invasion and metastasis. Furthermore, increased fascin

expression has been observed in tumors with more

aggressive clinical behavior including higher tumor stage,

lymph nodal stage, and clinical stage.18 As it relates to

invasive carcinomas of the urinary bladder, recent studies

have demonstrated an association with fascin expression

and aggressive urothelial carcinomas, with increased fas-

cin expression noted in microinvasive foci of T1 papillary

carcinomas.7,8 As demonstrated in previous studies, we

also found that fascin expression in normal urothelium

was absent to sporadic with intense immunoreactivity

noted in neoplastic urothelium, both in invasive and non-

invasive areas. These findings further support previous

observations of increased fascin expression in aggressive

urothelial carcinomas.

With the advent and increased utilization of liquid-based

cytology, we sought to characterize the utility of fascin

immunohistochemistry on SurePathTM urine cytology prepa-

rations in addition to biopsy specimens. Recent studies have

compared the SurePathTM technique with other conventional

MCKNIGHT ET AL.



methods of urine cytology preparation and found that Sure-

PathTM yields a greater overall cellularity, a more pro-

nounced tumor cell clustering, and a cleaner background

with fewer obscuring factors.19 Our results confirm the

extent of malignant tumor cell clustering observed previ-

ously, with fascin immunohistochemistry serving to further

Fig. 2. A: High-grade urothelial carcinoma invading muscularis propria with intense staining in tumor cells (Hematoxylin counterstain 310). B: Fascin reac-tivity in vascular endothelial cells and malignant cells in lymphovascular spaces (Hematoxylin counterstain 320). C: A focus of microinvasion in high-gradeurothelial carcinoma (arrow). Note absence of fascin reactivity in overlying carcinoma in situ. Corresponding urine cytology demonstrated fascin reactivity inmalignant cells (Hematoxylin counterstain320). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

UROTHELIAL CARCINOMA

Diagnostic Cytopathology, Vol 39, No 9 639


highlight these areas and enhance visualization. Addition-

ally, as immunohistochemistry is sometimes utilized on liq-

uid based preparations from various body sites, our study

demonstrates relevance with regard to fascin immunohisto-

chemistry and liquid based urine cytology, and it provides a

platform for the potential investigation of fascin expression

with regard to neoplasms in cytologic preparations from

other body sites.

Despite advances in urine cytology, the assessment of

invasion in tissue biopsies remains the gold standard in the

diagnosis of invasive urothelial carcinoma. Not only have

statistically significant differences been noted between

patients with invasive versus noninvasive tumors, there are

significant treatment differences between patients with inva-

sive and noninvasive tumors.12 Patients with tumor infiltra-

tion into the muscularis propria and beyond are generally

deemed candidates for more aggressive therapy with radical

cystectomy and possible neoadjuvant chemotherapy, both of

which are associated with significant morbidity. At the cur-

rent time there are no reliable methods to discern between

lamina propria and muscularis propria invasion outside of

tissue biopsy. Tissue biopsy has limitations as well, however,

and it is sometimes difficult to assess invasion on biopsy.

This has provided the impetus for investigations into novel

markers to aid standard diagnostic measures.

To our knowledge, this study is the first to utilize fas-

cin immunohistochemistry in urine cytology with biopsy

correlation. In conclusion, our results indicate that fascin

staining can be successfully performed on SurePathTM cy-

tology preparations. Fascin, while not a definitive marker

for invasion, is overexpressed in urothelial carcinomas of

the urinary bladder. Furthermore, intense fascin staining

in urine cytology specimens appears to be highly sensitive

and specific for urothelial carcinoma and correlates with

invasion on subsequent biopsy. Of particular interest is a

subset of biopsy cases in which fascin highlighted only

isolated nests of invasive tumor cells, and intense fascin

staining was noted in the concurrent urine cytology.

These findings suggest that fascin staining and utilization

in urine cytology might serve as a useful adjunct in pre-

dicting invasiveness in subsequent biopsies.

References1. Eble, JN, Sauter G, Epstein JI, et al. World Health Organization

classification of tumors. Pathology and genetics of tumors of theurinary system and male genital organs. Lyon: IARC Press; 2005.

2. Droller MJ. Bladder cancer: State-of-the-art care. CA Cancer J Clin

1998;48:269–284.

3. Bellmunt J, Albiol S, Kataja V, On behalf of the ESMO Guidelines

Working Group. Invasive bladder cancer: ESMO clinical recom-mendations for diagnosis, treatment and follow-up. Ann Oncol2009;20:iv79–iv80.

4. Jimenez RE, Gheiler E, Oskanian P, et al. Grading the invasive

component of urothelial carcinoma of the bladder and its relation-ship with progression-free survival. Am J Surg Pathol 2000;24:980–987.

5. Thieblemont C, Fendler JP, Trillet-Lenoir V, et al. Prognostic fac-

tors of survival in infiltrating urothelial bladder carcinoma. A retro-spective study of 158 patients treated by radical cystectomy. BullCancer 1996;83:139–146.

6. Hashimoto Y, Skacel M, Adams JC. Roles of fascin in human carci-

noma motility and signaling: Prospects for a novel biomarker? Int JBiochem Cell Biol 2005;37:1787–1804.

7. Tong G-X, Yee H, Chiriboga L, Hernandez O, Waisman J. Fascin-1

expression in papillary and invasive urothelial carcinomas of theurinary bladder. Hum Pathol 2005;36:741–746.

8. Karasavvidou F, Barbanis S, Pappa D, et al. Fascin determination in

urothelial carcinomas of the urinary bladder. Arch Pathol Lab Med2008;132:1912–1915.

9. Bassi PF, De Marco V, De Lisa A, et al. Non-invasive diagnostic

tests for bladder cancer: A review of the literature. Urologia Inter-nationalis 2005;75:193–200.

10. Quek ML, Quinn DI, Daneshmand S, Stein JP. Molecular prognosti-cation in bladder cancer—A current perspective. Eur J Cancer2003;39:1501–1510.

11. Vrooman OP, Witjes JA. Molecular markers for detection, surveil-lance and prognostication of bladder cancer. Int J Urol2009;16:234–243.

12. Sternberg SS, Mills SE, Carter D. Sternberg’s diagnostic surgicalpathology.5th ed. Philadelphia: Wolters Kluwer Lippincott Williams& Wilkins; 2010.

13. Piaton E, Faynel J, Hutin K, Ranchin M-C, Cottier M. Conventionalliquid-based techniques versus Cytyc Thinprep processing of urinarysamples: A qualitative approach. BMC Clin Pathol 2005;5:1–7.

14. Yamashiro S, Yamakita Y, Ono S, Matsumura F. Fascin, an actin-bundling protein induces membrane protrusions and increases cellmotility of epithelial cells. Mol Biol Cell 1998;9:993–1006.

15. Bakshi NA, Finn WG, Schnitzer B, Valdez R, Ross CW. Fascinexpression in diffuse large B-cell lymphoma. Anaplastic large celllymphoma, and classical hodgkin lymphoma. Arch Pathol Lab Med2007;131:742–747.

16. Durmaz A, Kurt B, Ongoru O, Karahatay S, Gerek M, Yalc NS.Significance of fascin expression in laryngeal squamous cell carci-noma. J Laryngol Otol 2010;124:194–198.

17. Won KY, Kim GY, Lim S-J, Park Y-K, Kim YW. Prognostic sig-nificance of fascin expression in extrahepatic bile duct carcinomas.Pathol Res Practice 2009;205:742–748.

18. Zou J, Yang H, Chen F, et al. Prognostic significance of fascin-1and E-cadherin expression in laryngeal squamous cell carcinoma.Eur J Cancer Prev 2010;19:11–17.

19. Lu DY, Nassar A, Siddiqui MT. High-grade urothelial carcinoma:Comparison of SurePath liquid-based processing with cytospin proc-essing. Diagn Cytopathol 2009;37:16–20.

MCKNIGHT ET AL.



fascin stain as a potential marker of invasiveness in carcinomas of the urinary bladder: a...

Documents