Aberrant Expression of p63 in Adenocarcinoma of the Prostate · Aberrant Expression of p63 in...

6
Aberrant Expression of p63 in Adenocarcinoma of the Prostate A Radical Prostatectomy Study Giovanna A. Giannico, MD,* Hillary M. Ross, MD,w Tamara Lotan, MD,w and Jonathan I. Epstein, MDwzy Abstract: Prostatic adenocarcinoma with aberrant diffuse ex- pression of p63 (p63-PCa) is a recently described variant of pro- static adenocarcinoma. The aim of this study was to investigate the clinical and pathologic features of p63-PCa at radical pros- tatectomy (RP). We reviewed 21 cases of p63-PCa diagnosed on needle biopsy at subsequent RP. Immunohistochemical analysis for PIN4 and Ki-67 was performed in all RP cases. p63-PCa showed a distinctive morphology consisting of atrophic, poorly formed glands, with multilayered and often spindled nuclei. Gleason grading was 3+3 = 6 in 28.5%, 3+5 = 8 in 38%, 3+4 = 7 in 14.3%, and 4+3 = 7, 5+3 = 8, and 5+4 = 9 in 9.5%. Usual-type acinar carcinoma coexisted in 85.7% with only p63-PCa present in the remaining cases. The usual-type carcinoma was Gleason grade 3+2 = 5 in 4.7%, 3+3 = 6 in 57%, 3+4 = 7 in 19%, and 4+3 = 7 in 4.3%. Overall, p63-PCa represented 65% of the total cancer volume (median 80%). The tumor was organ- confined in 16 cases (76.2%). In the remaining 5 cases, 2 had p63- PCa extending to the margin in areas of intraprostatic incisions, 2 had usual-type acinar adenocarcinoma extending to the margin and extraprostatic tissue, respectively, and 1 had p63-PCa with an unusual cribriform morphology involving the bladder neck. Ki-67 was low, <5% in all cases of p63-PCa, with similar expression in the coexisting acinar-type carcinoma. In summary, it is recom- mended that these tumors not be assigned a Gleason score and their favorable findings at RP be noted. Key Words: p63, radical prostatectomy, prostate cancer (Am J Surg Pathol 2013;37:1401–1406) P rostatic carcinoma with aberrant p63 expression (p63- PCa) is a recently described unusual variant of pros- tate cancer, characterized by distinct p63 expression of the secretory cells in a nonbasal distribution and lack of basal cell stain for high–molecular weight cytokeratin (HMWCK). Preliminary studies have identified lack of ERG gene rearrangement 1 and a mixed luminal and basal phenotype. 2 However, the significance of this phenom- enon and the biological behavior of this neoplasm are yet unknown. A previous study 3 described the features of this neoplasm predominantly in needle biopsy specimens. We undertook this study to further correlate the morphologic features with the biological behavior of this unusual neoplasm in radical prostatectomy (RP) specimens. MATERIALS AND METHODS Cases of p63-PCa diagnosed on needle biopsy at Johns Hopkins Hospital from 2008 to 2012 were reviewed; from these, 27 men who had undergone RP were selected. In all cases, the initial needle biopsies had been sent for consultation from other institutions. One of these men came for RP at Johns Hopkins Hospital. For all other cases, the RP slides were requested and obtained from the outside institutions, where the RP specimens were not processed in a uniform manner. For light microscopic evaluation, sections stained with hematoxylin and eosin were reviewed, and the following parameters were eval- uated: p63-PCa grade, usual-type adenocarcinoma grade, merging of different adenocarcinoma patterns versus for- mation of separate nodules, tumor volume, presence or absence of adjacent high-grade prostatic intraepithelial neoplasia (PIN), location [transition vs. peripheral zone (PZ)], organ-confined disease versus extraprostatic ex- tension, margin status, and lymph node metastasis. In addition, cytologic features, including cellular, nuclear, and nucleolar morphology were analyzed. Immunohistochemical analysis for p63, HMWCK, and a-methyl acyl coenzyme-A racemase (AMACR) in a predilute PIN4 Cocktail (AMACR+HMWCK+p63) from Biocare Medical (Concord, CA) and MIB-1 (mon- oclonal, mouse; Ventana; predilute) was performed in all RP cases at Johns Hopkins Hospital using 4-mm-thick sections obtained from formalin-fixed paraffin-embedded tissue. Immunohistochemical analyses were performed on a Ventana BenchMark XT automated stainer (Ventana, Tucson, AZ), including appropriate positive and negative control analyses. Ki-67-labeling index was measured on From the *Department of Pathology, Microbiology and Immunology, Vanderbilt University Medical Center, Nashville, TN; Departments of wPathology; zUrology; and yOncology, The Johns Hopkins Medical Institutions, Baltimore, MD. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Jonathan I. Epstein, MD, Department of Pathology, The Johns Hopkins Hospital, Weinberg 2242, 401 North Broadway, Baltimore, MD 21231 (e-mail: [email protected]). Copyright r 2013 by Lippincott Williams & Wilkins ORIGINAL ARTICLE Am J Surg Pathol Volume 37, Number 9, September 2013 www.ajsp.com | 1401

Transcript of Aberrant Expression of p63 in Adenocarcinoma of the Prostate · Aberrant Expression of p63 in...

Aberrant Expression of p63 in Adenocarcinoma ofthe Prostate

A Radical Prostatectomy Study

Giovanna A. Giannico, MD,* Hillary M. Ross, MD,w Tamara Lotan, MD,w andJonathan I. Epstein, MDwzy

Abstract: Prostatic adenocarcinoma with aberrant diffuse ex-

pression of p63 (p63-PCa) is a recently described variant of pro-

static adenocarcinoma. The aim of this study was to investigate

the clinical and pathologic features of p63-PCa at radical pros-

tatectomy (RP). We reviewed 21 cases of p63-PCa diagnosed on

needle biopsy at subsequent RP. Immunohistochemical analysis

for PIN4 and Ki-67 was performed in all RP cases. p63-PCa

showed a distinctive morphology consisting of atrophic, poorly

formed glands, with multilayered and often spindled nuclei.

Gleason grading was 3+3=6 in 28.5%, 3+5=8 in 38%,

3+4=7 in 14.3%, and 4+3=7, 5+3=8, and 5+4=9 in

9.5%. Usual-type acinar carcinoma coexisted in 85.7% with only

p63-PCa present in the remaining cases. The usual-type carcinoma

was Gleason grade 3+2=5 in 4.7%, 3+3=6 in 57%, 3+4=7

in 19%, and 4+3=7 in 4.3%. Overall, p63-PCa represented 65%

of the total cancer volume (median 80%). The tumor was organ-

confined in 16 cases (76.2%). In the remaining 5 cases, 2 had p63-

PCa extending to the margin in areas of intraprostatic incisions,

2 had usual-type acinar adenocarcinoma extending to the margin

and extraprostatic tissue, respectively, and 1 had p63-PCa with an

unusual cribriform morphology involving the bladder neck. Ki-67

was low, <5% in all cases of p63-PCa, with similar expression in

the coexisting acinar-type carcinoma. In summary, it is recom-

mended that these tumors not be assigned a Gleason score and

their favorable findings at RP be noted.

Key Words: p63, radical prostatectomy, prostate cancer

(Am J Surg Pathol 2013;37:1401–1406)

Prostatic carcinoma with aberrant p63 expression (p63-PCa) is a recently described unusual variant of pros-

tate cancer, characterized by distinct p63 expression of

the secretory cells in a nonbasal distribution and lack ofbasal cell stain for high–molecular weight cytokeratin(HMWCK). Preliminary studies have identified lack ofERG gene rearrangement1 and a mixed luminal and basalphenotype.2 However, the significance of this phenom-enon and the biological behavior of this neoplasm are yetunknown. A previous study3 described the features of thisneoplasm predominantly in needle biopsy specimens. Weundertook this study to further correlate the morphologicfeatures with the biological behavior of this unusualneoplasm in radical prostatectomy (RP) specimens.

MATERIALS AND METHODSCases of p63-PCa diagnosed on needle biopsy at

Johns Hopkins Hospital from 2008 to 2012 were reviewed;from these, 27 men who had undergone RP were selected.In all cases, the initial needle biopsies had been sent forconsultation from other institutions. One of these mencame for RP at Johns Hopkins Hospital. For all othercases, the RP slides were requested and obtained from theoutside institutions, where the RP specimens were notprocessed in a uniform manner. For light microscopicevaluation, sections stained with hematoxylin and eosinwere reviewed, and the following parameters were eval-uated: p63-PCa grade, usual-type adenocarcinoma grade,merging of different adenocarcinoma patterns versus for-mation of separate nodules, tumor volume, presence orabsence of adjacent high-grade prostatic intraepithelialneoplasia (PIN), location [transition vs. peripheral zone(PZ)], organ-confined disease versus extraprostatic ex-tension, margin status, and lymph node metastasis. Inaddition, cytologic features, including cellular, nuclear,and nucleolar morphology were analyzed.

Immunohistochemical analysis for p63, HMWCK,and a-methyl acyl coenzyme-A racemase (AMACR) in apredilute PIN4 Cocktail (AMACR+HMWCK+p63)from Biocare Medical (Concord, CA) and MIB-1 (mon-oclonal, mouse; Ventana; predilute) was performed in allRP cases at Johns Hopkins Hospital using 4-mm-thicksections obtained from formalin-fixed paraffin-embeddedtissue. Immunohistochemical analyses were performed ona Ventana BenchMark XT automated stainer (Ventana,Tucson, AZ), including appropriate positive and negativecontrol analyses. Ki-67-labeling index was measured on

From the *Department of Pathology, Microbiology and Immunology,Vanderbilt University Medical Center, Nashville, TN; Departmentsof wPathology; zUrology; and yOncology, The Johns HopkinsMedical Institutions, Baltimore, MD.

Conflicts of Interest and Source of Funding: The authors have disclosedthat they have no significant relationships with, or financial interestin, any commercial companies pertaining to this article.

Correspondence: Jonathan I. Epstein, MD, Department of Pathology,The Johns Hopkins Hospital, Weinberg 2242, 401 North Broadway,Baltimore, MD 21231 (e-mail: [email protected]).

Copyright r 2013 by Lippincott Williams & Wilkins

ORIGINAL ARTICLE

Am J Surg Pathol � Volume 37, Number 9, September 2013 www.ajsp.com | 1401

the basis of the proportion of tumor-positive cells usingMIB-1 proliferation marker.

RESULTSThe patients’ mean age was 59 years (range, 47 to

70 y). Of the 27 cases reviewed, 6 cases were excludedfrom the study: 2 cases due to absence of residual canceron RP, 1 case due to hormone therapy effect, 2 cases dueto absence of identifiable p63-PCa (only residual usual-type acinar carcinoma), and 1 case due to intense p63cytoplasmic and nuclear staining, the significance ofwhich was uncertain.

Of the remaining 21 cases, 18 (85.7%) p63-PCas hada distinctive morphology consisting of a mixture ofatrophic glands, some poorly formed or solid, with glandshaving multilayered, spindled, and basaloid nuclei(Figs. 1A, B). Nucleoli were present and prominent in themajority of the neoplastic glands. In the remaining 3(14.3%) cases, the neoplastic glands had no distinctivemorphology and were more akin to usual-type atrophicadenocarcinoma (Figs. 1C, D) without nuclear palisadingor basaloid morphology or showed neoplastic glandslined by columnar cells (Figs. 1E, F). In these cases, thediagnosis of p63-PCa was supported by im-munohistochemistry, which highlighted the distinctivepattern of nuclear positivity in the neoplastic glands.

p63-PCa was graded as 3+3=6 in 6 cases (28.5%),3+5=8 in 8 cases (38.1%), 3+4=7 in 3 cases (14.3%),4+3=7 or 5+3=8 in 2 cases (9.5%), and 5+4=9 in 2cases (9.5%) (Figs. 2A, B). Usual-type acinar prostatecarcinoma coexisted in 18 cases (85.7%) with only p63-PCa present in the remaining 3 cases. The usual-typecarcinoma Gleason grade was 3+2=5 in 1 case (4.7%),3+3=6 in 12 cases (57.1%), 3+4=7 in 4 cases(19.0%), and 4+3=7 in 1 case (4.3%). In the 21 cases,the mean of the total tumor volume occupied by p63-PCawas 65% (median 80%). In 14 of 18 cases in which acinar-type and p63-PCa coexisted, the latter was present inseparate nodules, whereas in the 4 remaining cases it wasadmixed with usual-type carcinoma (Figs. 2C, D). Thetumor was present in the PZ in 16 cases (76.2%), ex-tended from PZ to the transition zone in 1 case, and wasadjacent to high-grade PIN in 10 cases (47.6%). The tu-mor was organ-confined in 16 cases (76.2%). Among theremaining 5 cases, 2 had p63-PCa extending to the marginin areas of intraprostatic incisions, 2 had usual-type aci-nar adenocarcinoma extending to the margin and ex-traprostatic tissue, respectively, and 1 had p63-PCa withan unusual cribriform morphology involving the bladderneck. There were no lymph node metastases in all 12 of 21patients examined who received a lymph node dissection.

Immunohistochemistry supported the presence ofpositive nuclear staining for p63 in the neoplastic glandsin all cases. HMWCK was negative for basal cell stainingand showed focal cytoplasmic positivity in neoplastic cellsin 1 case (Figs. 2E, F). Two cases with distinct p63 nu-clear staining also showed positive cytoplasmic stainingfor AMACR. Ki-67 was low, <5% in all cases of p63-

PCa, with similar expression in the coexisting acinar-typecarcinoma (Fig. 3).

DISCUSSIONP63-PCa is a recently described3 unusual variant of

prostatic adenocarcinoma, characterized by aberrant ex-pression of p63 by immunohistochemistry and lack ofERG fusion.1 Its incidence is unknown. However, only 17cases of p63-PCa per year were diagnosed on needle bi-opsy in the consult service of one of the authors’ (J.I.E.),for a frequency of 0.3% of prostate cancers seen in hisconsultation practice. This frequency is an over-representation of its incidence, as consult cases would beenriched with unusual cancers. In contrast, p63-positivecancers may be more common than we realize in that 3/21cases analyzed in this study at RP did not display dis-tinctive morphology and were only detected by im-munohistochemical staining as a result of their needlebiopsy showing p63-PCa. As currently, there is no knownprognostic significance of p63-positive prostate cancer, itis not necessary for pathologists to routinely performimmunohistochemical analysis on biopsy or RP speci-mens in the absence of distinctive morphology.

The current study represents the first series of RPdescribing p63-PCa published to date. In the majority ofthe cases reviewed in the current study, the morphologyof this neoplasm was distinctive and easily recognizable.However, in a minority of cases a morphologic diagnosisof p63-PCa based solely on hematoxylin and eosinstaining could not be made, as the tumor resembledusual-type acinar atrophic adenocarcinoma or was linedby columnar cells. In such cases, immunohistochemicalanalysis for basal cell markers was crucial in recognizingthis entity. These cases without distinctive spindled mul-tilayered morphology consisting of predominantly atro-phic glands may represent a pitfall in needle biopsyspecimens, in which the infiltrative nature of the neoplasmmay not be readily apparent, mimicking benign atrophy.Therefore, in morphologically equivocal cases, im-munohistochemical analysis for basal cell markers shouldbe performed. p63-PCa is readily recognized when p63and HMWCK stains are done individually, with the latternegative for basal cells. However, in cocktails combiningp63 and HMWCK, we observed that it is more difficultfor contributors to note that only the nuclei are stainingfor p63 with an absence of cytoplasmic HMWCK. An-other difficulty with the interpretation of p63-PCa is itsdistinction from basal cell carcinoma. Before the recog-nition of p63-PCa, we had erroneously reported one ofthese cases as basal cell carcinoma. None of the patternsof basal cell carcinoma have the distinctive morphologyof p63-PCa. Basal cell carcinomas also are immunor-eactive with HMWCK, although often the more luminalcells are negative. The presence of well-formed and poorlyformed glands with p63 positivity could also bear super-ficial resemblance to sclerosing adenosis, but the lesionlacked the characteristic stromal cellularity, periglandular

Giannico et al Am J Surg Pathol � Volume 37, Number 9, September 2013

1402 | www.ajsp.com r 2013 Lippincott Williams & Wilkins

FIGURE 1. A, p63-PCa (PIN4 stain) with distinctive morphology of small atrophic glands, many cut tangentially, with basaloidfeatures with multilayered and spindled nuclei. B, Strong nuclear p63 positivity of case seen in (A). C, Atrophic adenocarcinomawith nondistinctive morphology showing perineural invasion (lower left). D, Same case as seen in (C) with diffuse p63 positivityand negative HMWCK (PIN4 stain). E, Unusual p63-PCa with less intense p63 immunoreactivity with columnar nuclei. F, Samecase as seen in (E) with p63 immunoreactivity and negative HMWCK (PIN4 stain).

Am J Surg Pathol � Volume 37, Number 9, September 2013 Prostate Cancer Aberrant p63 Expression

r 2013 Lippincott Williams & Wilkins www.ajsp.com | 1403

FIGURE 2. A, p63-PCa that is 5+5 = 10 according to the Gleason grading system. Inset shows p63-PCa. Note benign gland (lowerleft) where basal cells are positive for both nuclear p63 and cytoplasmic HMWCK (PIN4 stain). B, p63-PCa that is 3+4 = 7according to the Gleason grading system. C, p63-PCa admixed with usual-type prostate adenocarcinoma. D, Im-munohistochemistry highlights the 2 different components in which the usual adenocarcinoma glands are negative for p63 (PIN4stain). E, p63-PCa with separate sections labeled with HMWCK (left panel) and HMWCK/p63 (right panel). Small cancer glandsare negative for HMWCK (left panel) and express p63 (right panel). Larger benign glands (right panel) express both HMWCK andp63. F, p63-PCa with p63 and HMWCK expression infiltrating between benign glands. Inset higher magnification.

Giannico et al Am J Surg Pathol � Volume 37, Number 9, September 2013

1404 | www.ajsp.com r 2013 Lippincott Williams & Wilkins

hyaline sheaths, and HMWCK positivity in stroma andbasal cells surrounding glands.

Another issue concerns the application of theGleason scoring system to this neoplasm. Indeed, 15 ofour cases (71.3%) were graded with a Gleason score Z7,including 12 (57%) that were graded Z8. However, de-spite the grading applied as such, the majority of thesecases were organ-confined (pT2), underscoring the dis-crepancy between the application of the Gleason grad-ing and biological behavior of this tumor. This arguesthat Gleason grading system may not apply for thesetumors.

The significance of p63 expression in this unusualvariant of prostatic adenocarcinoma is yet unknown. p63is a member of the p53 family, and it is expressed in 2isoforms with opposing tumor suppressor and oncogenicfunctions: Tap63, a full-length form, predominantly ex-pressed in oocytes; and DNp63, an amino-deleted form,expressed in the epidermis, where it plays a role in epi-thelial development. Mice homozygous for a disruptedp63 gene have lethal absence of the epidermis and a widerange of other epithelia, including prostate.4,5 In adult-

hood, p63 acquires a role in the process of tumorigenesis,and it is expressed in neoplastic cancers of the head andneck, thyroid, prostate, breast, and lung, where it may beinvolved in the persistence of cancer stem cells,6 epithelialmesenchymal transition,7,8 senescence,9,10 and apoptosis.11

Preliminary studies have shown that the DNp63,detected by the p40 antibody, was expressed in 15/16(94%) cases of p63-positive prostate adenocarcinomas,12

whereas CK18, androgen receptor, and AMACR wereexpressed in 100%, 90%, and 92%, respectively, withonly focal weak staining for basal cytokeratin 5/6.2 Inaddition, a previous study3 showed Bcl2 positivity in71.4% of cases analyzed. Therefore, a mixed luminal-basal phenotype has been proposed for this neoplasm. Wecould further speculate an origin from a pluripotentDNp63-positive progenitor that has an ability not only tomaintain the normal epithelial stem population but alsohas a role in the persistence of cancer stem cells; therefore,the persistence of p63 expression would drive the onco-genic property of this neoplasm. It has also been pro-posed that p63 loss accelerates tumorigenesis andmetastatic spread, suggesting that p63-positive cancers

FIGURE 3. Same radical prostatectomy with (A) usual-type acinar adenocarcinoma, (B) Ki-67 expression in usual-type acinaradenocarcinoma, (C) p63-PCa with distinctive morphology, (D) Ki-67 expression in p63-PCa acinar adenocarcinoma.

Am J Surg Pathol � Volume 37, Number 9, September 2013 Prostate Cancer Aberrant p63 Expression

r 2013 Lippincott Williams & Wilkins www.ajsp.com | 1405

may be associated in some circumstances with less ag-gressive behavior.13,14 This was supported in our study bythe finding that almost all cases were confined to theprostate, by the findings of low Ki-67 expression, and bythe absence of metastatic disease in the cohort analyzed.Proteomic and immunohistochemical cluster analysis ofprostate tissue showed that p63, among other genes wasoverexpressed in localized and underexpressed in meta-static disease.15 Similar results were obtained in awatchful waiting Swedish cohort of patients with lo-calized prostate cancer.16 It is of interest that cytoplasmicexpression of p63 in prostate tumor tissue was associatedwith an increased proportion of cancer deaths among298 men, with increasing risk in cases with higher in-tensity of cytoplasmic p63 expression in a study. It hasbeen proposed that the predominant cytoplasmicrather that nuclear localization could represent abnormalnuclear-cytoplasmic shuttling and disrupted cell cycleregulation.17

Despite representing a relatively rare phenomenon,recognition of this unusual pattern is important for sev-eral reasons. It is critical to recognize the neoplastic na-ture of this cancer variant, whose pattern may beinterpreted as benign or atypical when p63/HMWCKbasal cell cocktails are used and when only nuclear p63staining is present. In addition, if graded according to theGleason scoring system based on glandular morphologyand architectural pattern, the majority of these lesionswould qualify for a high Gleason score (7 or higher) be-cause of the frequent presence of poorly formed glandularmorphology or single cells, potentially leading to over-treatment or exclusion from active surveillance protocols.It is thus critical to recognize these unique cancers andrecommended that these tumors not be assigned a Glea-son score and their favorable findings at RP be noted.Further follow-up is needed to fully characterize theirbiological behavior.

REFERENCES1. Baydar DE, Kulac I, Gurel B, et al. A case of prostatic

adenocarcinoma with aberrant p63 expression: presentation withdetailed immunohistochemical study and FISH analysis. Int J SurgPathol. 2011;19:131–136.

2. Lotan TL, Osunkoya AO, DeMarzo AM, et al. Prostate Adeno-carcinomas Aberrantly Expressing p63 Have a Luminal CellPhenotype. Poster presented at: The 101st Annual Meeting of theUnited States and Canadian Academy of Pathology (USCAP);-March 19 2012; Vancouver, Canada.

3. Osunkoya AO, Hansel DE, Sun X, et al. Aberrant diffuse expressionof p63 in adenocarcinoma of the prostate on needle biopsy andradical prostatectomy: report of 21 cases. Am J Surg Pathol.2008;32:461–467.

4. Yang A, Schweitzer R, Sun D, et al. p63 is essential for regenerativeproliferation in limb, craniofacial and epithelial development.Nature. 1999;398:714–718.

5. Mills AA, Zheng B, Wang XJ, et al. p63 is a p53 homologuerequired for limb and epidermal morphogenesis. Nature. 1999;398:708–713.

6. Senoo M, Pinto F, Crum CP, et al. p63 is essential for theproliferative potential of stem cells in stratified epithelia. Cell.2007;129:523–536.

7. Blanpain C, Fuchs E. Epidermal homeostasis: a balancing act ofstem cells in the skin. Nat Rev Mol Cell Biol. 2009;10:207–217.

8. Higashikawa K, Yoneda S, Tobiume K, et al. Snail-induced down-regulation of deltaNp63alpha acquires invasive phenotype of humansquamous cell carcinoma. Cancer Res. 2007;67:9207–9213.

9. Keyes WM, Wu Y, Vogel H, et al. p63 deficiency activates aprogram of cellular senescence and leads to accelerated aging. GenesDev. 2005;19:1986–1999.

10. Keyes WM, Wu Y, Vogel H, et al. p63 deficiency activates aprogram of cellular senescence and leads to accelerated aging. GenesDev. 2005;19:1986–1999.

11. Gressner O, Schilling T, Lorenz K, et al. TAp63a induces apoptosisby activating signaling via death receptors and mitochondria.EMBO J. 2005;24:2458–2471.

12. Uchida K, Ross HM, Epstein JI, et al. DeltaNp63 Isoforms of p63in Aberrant Diffuse p63 Positive Prostate Cancer. Poster presentedat: The 101st Annual Meeting of the United States and CanadianAcademy of Pathology (USCAP); March 19, 2012; Vancouver,Canada.

13. Urist MJ, Di Como CJ, Lu ML, et al. Loss of p63 expression isassociated with tumor progression in bladder cancer. Am J Pathol.2002;161:1199–1206.

14. Koga F, Kawakami S, Fujii Y, et al. Impaired p63 expressionassociates with poor prognosis and uroplakin III expression ininvasive urothelial carcinoma of the bladder. Clin Cancer Res.2003;9:5501–5507.

15. Bismar TA, Demichelis F, Riva A, et al. Defining aggressiveprostate cancer using a 12-gene model. Neoplasia. 2006;8:59–68.

16. Mucci LA, Pawitan Y, Demichelis F, et al. Testing a multigenesignature of prostate cancer death in the Swedish Watchful WaitingCohort. Cancer Epidemiol Biomarkers Prev. 2008;17:1682–1688.

17. Dhillon PK, Barry M, Stampfer MJ, et al. Aberrant cytoplasmicexpression of p63 and prostate cancer mortality. Cancer EpidemiolBiomarkers Prev. 2009;18:595–600.

Giannico et al Am J Surg Pathol � Volume 37, Number 9, September 2013

1406 | www.ajsp.com r 2013 Lippincott Williams & Wilkins