[CANCER RESEARCH 58. 542-548. February 1. 1998)

Loss of Cyclin-dependent Kinase Inhibitor p27Klpl Is a Novel Prognostic Factor in

Localized Human Prostate Adenocarcinoma1

John Tsihlias,2 Linda R. Kapusta,2"3 Gerrit DeBoer, Izabella Morava-Protzner, Ingrid Zbieranowski,Nandita Bhattacharya, G. Charles Catzavelos, Laurence H. Klotz, and Joyce M. Slingerland3

Department of Pathology [L R. K.. I. M-P., G. C. C.I, Cancer Biology Research ¡J.T., N. B.. J. M. S.], and Division of Urology []. T.. L H. K.¡.Sunnyhrook Health ScienceCenter: Department of Politologi: Women's College Hospital ¡I.Z.]; and Departments of Medical Oncology ¡J.M. SJ and Clinical Trials and Epidemiology ¡G.D./. Toronto-

Sunnvbrook Regional Cancer Center. University of Toronto, Toronto, Ontario, M4N 3M5, Canada

ABSTRACT

p27Klpl is a cyclin-dependent kinase inhibitor that negatively regulates

cell proliferation by mediating cell cycle arrest in G,. This study wasundertaken to assess the prognostic value of p27Klpl in localized human

prostate cancer. Archival material from 113 radical prostatectomy specimens obtained between 1985 and 1993 was stained immunohistochemi-cally for p27Kipl protein using a commercially available antibody. Patient

charts were reviewed for preoperative serum prostate-specific antigen,

clinical and pathological staging, Gleason tumor grade, time to biochemical and clinical recurrence, and survival. Strong p27Klpl staining was

uniformly seen in benign prostatic epithelial components in all tumorsections. p27Kipl staining was reduced in most prostate cancers and wasvariable in prostatic intraepithelial neoplasia. Decreased p27Kipl staining«25'r of nuclei stained positive for p27KIpl) correlated with seminalvesicle involvement (/' = 0.0032) and with higher Gleason gradeI/' = 0.0114). On univariate analysis, low p27Kipl predicted an increasedrisk of treatment failure in the node-negative cohort (/' = 0.0037) and inthe subset who did not receive neoadjuvant hormonal therapy (/' = 0.049).Low p27Klpl expression was an independent predictor of treatment failure

on multivariate analysis of lymph node negative prostate cancers following radical retropubic prostatectomy (n = 102; /" = 0.047). Seminal vesicleinvolvement (/' = 0.034) and positive surgical margins (/' = 0.047) were

also independent prognostic factors for disease recurrence. In patientswho received preoperative neoadjuvant hormonal therapy, low p27Klpl in

the pathological specimen was an even stronger predictor of outcome thanit was in the entire group (n = 23, P = 0.015).

INTRODUCTION

Prostate carcinoma is the most common solid malignancy and thesecond most common cause of cancer-related death in men. There

were 337,000 estimated new cases of prostate cancer and 45,400estimated prostate cancer-related deaths in North America in 1996 (1,2). Many of these cancers are slow-growing, well-differentiated tu

mors that do not affect longevity. Others are rapidly growing andaggressive (3-5). Traditional parameters used to predict extent of

disease and natural history include: clinical stage, tumor grade, andserum PSA.4 Published nomograms allow more accurate prediction of

pathological stage using these three parameters (6). Most patients,however, fall in a middle zone where meaningful prognostication

Received 9/16/97; accepted 12/2/97.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked atlvertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was funded by the Canadian Breast Cancer Research Initiative. J. T. is a

Fellow of the National Cancer Institute of Canada, supported by funding from the TerryFox Run, and was also supported by the Toronto-Sunnybrook Regional Cancer Center andCancer Care Ontario. J. M. S. is a Clinician Investigator, supported by Cancer CareOntario.

2 The first two authors contributed equally to this work.-1Requests for reprints should be addressed to either Dr. L. Kapusta. Department of

Pathology, or Dr. J. Slingerland. Cancer Biology Research, Sunnybrook Health ScienceCenter, 2075 Bayview Avenue, Toronto, Ontario, Canada, M4N 3M5. Phone: (416) 480-4120 or (416) 480-6100. ext. 3494; Fax: (416)480-5703; E-mail:joyce.slingerland@utoronto.ca.

4 The abbreviations used are: PSA, prostate-specific antigen; cdk, cyclin dependent

kinase; INK4, inhibitor of cdk4; TNM. tumor-node-metastasis; CAS, Cell AnalysisSystems; NHT, neoadjuvant hormonal therapy; KIP, kinase inhibitor protein.

based on these factors is difficult. A more accurate prognostic markerfor this patient group would permit assignment of more aggressivetreatment to those most likely to fail and would justify expectantmanagement (deferral of active treatment) in those likely to haveindolent disease.

There is increasing evidence that cell cycle regulators are disruptedin human cancers (7). The cell cycle is governed by cdks, the activitiesof which are regulated by binding of positive effectors, the cyclins (8,9); by negative regulators, the cdk inhibitors (10); and by phospho-

rylation and dephosphorylation events (11, 12). The cdks integratemitogenic and growth-inhibitory signals and coordinate cell cycle

transitions (13, 14). Passage through G, into S phase is regulated bythe activities of cyclin D-, cyclin E-, and cyclin A-associated kinases.B-type cyclin-associated kinases regulate G2-M. Two families of cdkinhibitors have been identified. INK4 family members pl5INK4B,p!6INK4A, p 18, and pl9 bind cyclin D-dependent kinases, specifically,

cdk4 and cdko. Kinase inhibitor protein (KIP) family members, including p21cipl, p27Kipl, and p57Kip2, bind and inhibit their targets.

The cdk inhibitors regulate entry into and progression through the cellcycle by modulating the activity of cdks in response to mitogenic andantimitogenic stimuli. The cyclins, cdks, and cdk inhibitors are frequently altered in cancer or disrupted secondarily by other oncogenicevents (7).

p27Klpl is a cdk inhibitor that regulates progression from G, into S

phase by binding to the cyclin E-cdk2 complex and inhibiting this

kinase (15). Regulation of this protein appears to occur primarily atthe posttranslational level by ubiquitin-mediated degradation (16).p27Klpl binds and inhibits its target cdks in a stoichiometric manner.Thus, the availability of p27Klpl is critical in the regulation of cyclin/cdk activity. Although mutations in the p27Ktpl gene are rare inhuman tumors (17-20), decreased p27Klpl protein has potential im

portance as a prognostic factor in breast (21, 22), colon (23), andgastric carcinoma (24), and its reduction may be associated withtumor progression. Here, we demonstrate that the loss of p27Klpl

appears to be of prognostic value in primary resectable prostatecarcinoma.

PATIENTS AND METHODS

Patient Population. The study group consisted of 113 patients who hadundergone radical prostatectomy, performed by one of two surgeons, at Sunnybrook Health Science Center between January 1986 and April 1993 and forwhom clinical follow-up data were available. Paraffin blocks from the embed

ded radical prostatectomy specimens were selected on the basis of having bothbenign and malignant histology present within the same tumor section. Theclinical database prepared from patient records included preoperative PSA,clinical and pathological staging, Gleason tumor grade, time to treatmentfailure, and survival. The patients were staged according to the TNM system(25). Treatment failure was defined as any one of the following: institution ofhormonal therapy postoperatively. failure of the PSA to return to undetectablelevels postoperatively (<0.2 /xg/liter; Hybritech assay), or any new PSA

elevation above undetectable levels on two successive readings during postoperative follow-up. Patients who had lymph node metastasis at the time of

542

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27IÃœPI:A PROGNOSTIC FACTOR IN PROSTATE CANCER

surgery were treated routinely with postoperative (adjuvant) hormonal therapy.Because institution of hormone therapy is within our definition of treatmentfailure, node-positive eases were excluded from the prognostic factor analysis.

Immunohistochemistry. Expression of p27Klpl protein was determined

immunohistochemically. Paraffin sections of tumor blocks were deparaffinizedwith xylene. rehydrated, and microwaved for 10 min in 10 mM citrate buffer(pH 6.0). Sections were blocked with 3% hydrogen peroxide in methanolfollowed by normal horse serum (1:20 dilution) in 0.1 M PBS at pH 6.0 andthen incubated overnight at 4°Cwith anti-p27Kipl monoclonal antibody (Trans-

duction Laboratories, Lexington, KY) diluted 1:1000 (0.25 ng/ml) in PBS.Slides were then reacted with biotin-labeled anti-mouse IgG and incubatedwith preformed avidin-biotin-peroxidase complex (Vector Laboratories,Burlingame. CA). Metal-enhanced diaminobenzidine substrate (Pierce, Rock-

ford, IL) was then added in the presence of horseradish peroxidase. Sectionswere counterstained with hematoxylin, dehydrated, and mounted. The degreeof p27Klpl staining was scored independently by two pathologists (L. R. K. and

I. M-P.). All sections selected contained both benign and malignant histology,which allowed comparison of immmunohistochemical staining of p27Klpl in

benign and malignant tissue. Negative controls using isotype-specific poly-

clonal mouse immunoglobulin showed no staining. Furthermore, the pattern ofp27 staining in tumor samples obtained with the monoclonal antibody fromTransduction Laboratories was confirmed using three different polyclonalp27Klpl sera (one from Santa Cruz Biotechnology, Santa Cruz, CA; and two

kindly provided by T. Hunter, the Salk Institute. La Jolla. CA; Ref. 26). p27staining with the Santa Cruz antibody was lost when the antibody was pread-

sorbed with blocking p27 peptide (Ref. 21 and data not shown). Between 15and 20 high-power fields of tumor were scored for the percentage of nucleishowing positive p27Klpl staining. The scoring system used was as in our

previous studies (21): 1, 0-25%; 2, 25-50%; 3, 50-75%; and 4, >75%

positive nuclear staining.Image Analysis. Image cytometric analysis was performed with the CAS

200 (Becton Dickinson Cellular Imaging Systems, San Jose, CA), a video-

based interactive image cytometer (27). using the Quantitative Nuclear Antigen Program software (Version 3.06; CAS Inc., Becton Dickinson CellularImaging Systems). Four negative fields on each slide were analyzed to establish the nuclear and antibody threshold optical densities by which the imageanalyzer could discriminate positive nuclear staining from background. Toprovide a second independent method to verify nuclear p27Klpl staining scores,

image analysis was performed on selected cases scored previously by percentage of positive nuclei. Three cases in each of the p27Klpl staining categories

(categories 1-4) were studied by image analysis. Between 20 and 30 nuclei

from normal prostate epithelial cells and 50 nuclei from carcinoma cells wereselected randomly and outlined on the video screen. Total nuclear area,percentage positive nuclear area, and the sum of the optical densities were

quantified for each case, for both benign and malignant nuclei. A minimumtotal nuclear area of 5000 /¿m2was measured for each case by the CAS image

analyzer.Statistical Methods. Time to treatment failure and its dependence on

putative prognostic factors was investigated by the log-rank method (28) andCox's proportional hazards regression analysis (29). Stepwise selection of

variables was used to determine the best predictors. Kaplan-Meier curves were

constructed (30) to show the probability of remaining free of failure as afunction of time after diagnosis. Correlation analysis between prognosticfactors was done using Spearman and Pearson coefficients.

RESULTS

Clinical and Pathological Cohort Data. Median patient age was65 years (range, 39-75 years), and median follow-up was 4.6 years(range, 0.75-11.4). The clinical and pathological data are summarized

in Table 1. Gleason scores were stratified into three groups, scores2-6, score of 7, and scores 8-10, based on the observation that a

Gleason score of 7 confers a significantly worse prognosis than doesa score of 6 (31, 32). The preoperative Gleason score was known for87 patients; preoperative serum PSA was available in 96 patients. Allother data were available for the entire cohort (n = 113). The majorityof patients were clinical stage T2 (n = 93), and the remainder were T,(n = 20). Neoadjuvant hormonal therapy (NHT) had been adminis-

Table 1 Clinical and pathnloaical data

Parametern(%)Preoperative

factorsPreoperativePSA*<44-1010-20>20Clinical

T-stage'TibTicT2aT2bT2cNeedle

biopsy Gleasongrade2-678-10Neoadjuvant

hormonaltherapyYesNoPostoperative

factorsGleasongrade from resectedtumor2-678-10Extra

capsularextensionYesNoMargins

positiveYesNoSeminal

vesicleinvolvementYesNoLymph

nodeinvolvementYesNoCurrent

statusAliveDead

(cancer or causeunknown)Dead(other cause)Total

(%)113(100)18(19)40(42)26

(27)12(12)11

(10)9(8)56

(50)25(22)12(11)57(66)22(25)8(9)24(21)89

(79)49

(43)45(40)19(17)76

(67)37(33)70

(62)43(38)21

(19)92(81)11(10)102

(90)104(92)3(3)6(5)<25%

p27"22(19.5)26652368396422069718418410123192002&25%p2791

(80.5)1634207965017948164226943361258335239U808838434

1Less than 25% of nuclei stained positively for p27 lp on immunohistochemislry." In ng/liter.' Clinical T-stage per the TNN staging system (see text for details).

tered preoperatively to 24 of 113 patients in a variable manner. Thedrug regimens used were cyproterone acetate (14 patients), diethyl-stilbesterol (5 patients), leuprolide plus nilutamide (2 patients), finas-

teride (2 patients), and flutamide alone ( I patient). Duration of hormonal treatment was also variable, with median duration of 4 weeks,a mean of 6.1 weeks, and a range of 1-26 weeks. Overall survival was

similar to that observed in other contemporary studies on outcomes ofradical prostatectomy (33-35).

Immunohistochemistry Results and Image Analysis. Strong expression of p27K'pl was consistently noted in virtually all nuclei of

benign prostatic epithelium, and there was a consistent but variabledegree of reduction in both the intensity of p27K'pl staining and the

number of nuclei stained in prostate carcinoma (Fig. 1). Although thenuclei of prostatic intraepithelial neoplasia tended to be p27Klpl

positive, there was a variable reduction in intensity of staining.Image analysis provided a validation of the visual scoring system

used for quantitation of p27Klpl nuclear staining by immunohisto-

chemistry. Image analysis of benign nuclei consistently showed thehighest values of summated absorbance in all cases studied (n = 12;range, 144-186; mean, 154.4). Over 150 nuclei were evaluated ineach category of p27Klpl staining (percentage nuclear staining cate

gory 1, 2, 3, or 4). Of malignant tumor nuclei, cases scored as 1(<25% nuclei staining) had the lowest summated absorbance values(range, 40-46; mean, 42.6), whereas cases scored as 4 (>75% nu

clear staining) had the highest value for malignant nuclei (range,107-117; mean, 112.0). Cases in the intermediate range, scored as 2

543

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27Klpi: A PROGNOSTIC FACTOR IN PROSTATE CANCER

Fig. I. p27K'pl nuclear staining of prostale ad-enocarcinoma. Paraffin-embedded material from

radical prostatectomy specimens was stained forp27KlP' and counterstained with hematoxylin. The

percentage of malignant nuclei with positive nuclear staining was scored independently by twopathologists into four categories. Large arrowheads, benign prostatic glandular epithelium; smallarrows, prostatic adenocarcinoma. A. adenocarci-

noma of prostate, Gleason score 4. with no nuclearstaining for p27Klp', category 1 (0-25% nuclei

staining positive). B, adenocarcinoma of prostate.Gleason score 3, with p27Klpl nuclear staining cat

egory 2 (25-50% nuclei positive). C, adenocarcinoma of prostate. Gleason score 3. with p27Klpl

nuclear staining category 3 (50-75% nuclei posi

tive). D, adenocarcinoma of prostate, Gleason score3 with high nuclear p27Klpl staining, category 4

(>75% nuclear staining). Magnification (A-D),

X400.

K**»**^

~c-

%*A •¿�*'

A»

Table 2 Correlation of p27K'1'' with other prognostic parameters"

p27K"" n Correlation coefficient

All patients (n =113)Seminal

vesicleinvolvementGleasongrade from resectedtumorMargins

positivePreoperativePSA ' (Pearsoncoefficient)Extracapsular

extensionNHTLymph

nodeinvolvementAge(Pearsoncoefficient)Preoperative,

NHT excluded (n =89)Seminal

vesicleinvolvementGleasongrade from resectedtumorMargins

positivePreoperativePSA* (Pearsoncoefficient)Eixtracapsular

extensionLymphnodeinvolvementAge

(Pearson coefficient)113113113%11311311311389898973898989-0.340-0.202-0.201-0.198-0.153+0.146-0.065-0.055-0.309-0.267-0.179-0.157-0.156-0.064-0.0710.00020.0320.0330.0540.1070.1230.500.560.00320.01

140.0930.1850.1440.550.51

" Low p27Klpl, <25% nuclear staining; high p27Klpl, s25% nuclear stainin

man rank correlation coefficient, except where otherwise noted.* In fig/liter.

ig; Spear-

(25-50% nuclear staining) or 3 (50-75% nuclear staining), had sum-

mated absorbance values between those of groups 1 and 4: casesscored as 2 had a range of 58-71 and a mean of 63.0, and cases scoredas 3 had a range of 83-92 and a mean of 88.3.

Prognostic Role of p27Kipl. In a correlation analysis betweenprognostic factors, low 27Klpl was strongly associated with seminal

vesicle involvement and with increased Gleason grade (P = 0.0002and P = 0.032 respectively; Table 2). A trend toward increasing PSAwas observed when p27Kipl was low (P = 0.054). Low p27Kipl did

not correlate significantly with extracapsular extension, lymph nodeinvolvement, use of NHT, or patient age (Table 2).

Univariate analysis of lymph node-negative cases (n = 102) by Coxregression or by the log-rank method identified the following as

prognostic for risk of treatment failure: seminal vesicle involvement(P = 0.0030), low p27Kipl staining (<25% positive nuclei;

P = 0.0037), postoperative Gleason grade (0.0094), and positivesurgical margins (P = 0.030; Table 3A).

Table 3 Univariate analysis for time to treatment failure

Prognostic marker P"

Node-negative cases (n = 102)*

Seminal vesicle involvement 0.0030<25% nuclei p27 positive 0.0037Gleason grade from resected tumor (2-6, 7, 8-10) 0.0094

Margins positive 0.030Needle biopsy gleason grade (2-6. 7. 8-10; n = 78) 0.051

Neoadjuvant hormonal therapy 0.25Clinical T-stager 0.26rfPreoperative PSA' (<4, 4-10, 10-20, >20; n = 86) 0.35d

Extracapsular extension 0.38Age 0.42

Node-negative cases, excluding NHT (n = 79)Gleason grade from resected tumor (2-6, 7, 8-10) 0.0027Needle biopsy Gleason grade (2-6. 7. 8-10; n = 58) 0.012Seminal vesicle involvement 0.044<25% nuclei p27 positive 0.049Margins positive 0.059Extracapsular extension 0.30Preoperative PSA' (<4, 4-10, 10-20, >20; n = 64) O^Clinical T-stage' 0.65d

Age 0.74" From score )f or log-rank test.* Sample size, except when indicated otherwise.' Clinical T-stage per the TNM staging system (see text for details).J P for linear trend.' In

544

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27KIHI: A PROGNOSTIC FACTOR IN PROSTATE CANChK

01 234567 9 10 11 12

100 :

80

60

40 :

20p = 0.0037

25% p27

01 23456789 10 11 12

Years

100 -

80 ;

J60'S0)i1

40:g20

;1

'-m—«---h>25%p271

¡+H

H<25%p27p

= 0.049

01 23456789 10 11 12

Years

Fig. 2. A, Kaplan-Meier curve of time to treatment failure (n = 102). B, Kaplan-Meiercurves of time to treatment failure, comparing patients with <25% p27Klpl nuclearstaining to those with >25% p27Kipl nuclear staining (n = 102; P = 0.0037). C,

Kaplan-Meier curve in patients with node-negative cancers, excluding those who receivedNHT, comparing <25% p27Klpl nuclear staining to those with 225% p27Klpl nuclear

staining (n = 79; P = 0.049).

The Kaplan-Meier curve of time to treatment failure for node-negative study patients showed a 5-year actuarial failure-free rate of66% (±9.6%, 95% confidence interval; Fig. 14). When the disease-free rate was plotted versus p27Klpl staining scores of 1, 2, 3, or 4,

only those tumors showing low p27Klpl nuclear staining (<25%

nuclei positive) had a distinctly worse prognosis over the follow-upperiod than did the remainder of the cohort (P = 0.0037; Fig. 25,Kaplan-Meier curves). The probabilities of being free of treatmentfailure at 5 years were 38 ± 23% for low-p27Klpl tumors and72 ±10% for the high-p27Kipl tumors.

Multivariate analysis of node-negative tumors (n = 102) identifiedlow p27Klpl as a novel independent prognostic factor for time to

treatment failure (P = 0.047; risk ratio = 2.08; Table 4). Seminal

vesicle involvement and positive surgical margins were also of independent prognostic value, with risk ratios of 2.29 (P = 0.034) and2.07 (P = 0.047). respectively (Table 4). Confirmation of these

observations in other larger data sets is desirable because the 95%confidence intervals are broad.

Effects of NHT. The effect of NHT on time to treatment failure didnot reach statistical significance (P = 0.25; Table 3A). A trend towardincreased p27Klpl staining was observed in tumors from patients who

received NHT. Only 2 of 24 tumors (8.3%) resected following an-drogen ablation therapy had low p27Klpl (<25% positive nuclei),

whereas 20 of 89 tumors (22.5%) that did not receive NHT had lowp2yKipi yjjjj, (Jifferengg was nO( statistically significant (P = 0.15;two-tailed Fisher's exact test). NHT had a significant effect on the

distribution of grade in the resected specimens. Among the 89 patientswho did not receive NHT, Gleason scores of 2-6, 7, and 8-10 were

observed in 44 (49%), 34 (38%), and 11 (12%) patients, respectively.In the 24 patients who received NHT, 5 (21%) had Gleason scores of2-6, 11 (46%) had Gleason scores of 7, and 8 (33%) had Gleasonscores of 8-10. This trend toward increased Gleason grade followingNHT was statistically significant ()f for trend = 8.7; P = 0.003).

Interestingly, when node-negative cancers treated with preoperativeNHT were analyzed separately, low p27Klp' in the resected tumor was

an even more significant predictor of reduced time to treatment failurethan it was in the group as a whole (P = 0.015; n = 23). When the

patients who received NHT were removed from the analysis, lowp27Kipl still predicted reduced time to relapse (P = 0.049; Table 3B

and Fig. 2C, Kaplan-Meier curve). On multivariate analysis of thecases with negative lymph nodes who did not receive NHT (n = 79),

only Gleason grade proved to be a statistically significant prognosticfactor for relapse-free rate (/> = 0.0034; risk ratio = 2.10).

DISCUSSION

Established modalities for managing localized prostate cancer include radical prostatectomy, radical radiation therapy, and surveillance. Untreated low-grade cancers in men with life expectancies ofless than 10-15 years are not likely to affect longevity (3). When

radical therapy is not given to patients with tumors of moderate orhigh grade, there is a loss in life years of 4-5 and 6-8 years,

respectively (3). Radical prostatectomy has a low mortality but significant morbidity, with impotence resulting in 50-70% and stressincontinence in 2-6% of cases (34-36). In tumors treated by radical

Table 4 Multivariate analysis for lime to treatment failure

Model for treatment failure Risk ratio (CD"

Node-negative cases (n = 102)

Seminal vesicle involvement<25% p27 positive nucleiMargins positive

NHT recipients excluded (n = 79)

Gleason grade from resectedtumor (2-6. 7. 8-10)2.29

(1.07-4.92)2.08 (1.01-4.27)2.07(1.01-4.24)2.10''

(1.28-3.44)0.034

0.0470.0470.0034

" CI. 95% confidence interval for the relative risk ratio.h From Wald x~-' Per step.

545

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27lap': A PROGNOSTIC FACTOR IN PROSTATE CANCER

prostatectomy, increasing grade predicts a shorter disease-free survival (37). When the cancer is organ confined, however, disease-free

survival is excellent, regardless of tumor grade (38). Because radicalprostatectomy is unlikely to be curative in locally advanced disease, amarker that accurately predicts whether or not cancer is organ-con

fined would allow better treatment assignment.Low p27Klpl correlated with seminal vesicle involvement and with

positive surgical margins in the entire cohort but not with extracap-

sular extension. Of 19 tumors with poorly differentiated disease(Gleason sum of 8-10), 7 had low p27Kipl. Pathological staging

revealed that none of these seven were organ confined. Of the remaining 12 high-grade tumors without low p27Klpl, 3 were organconfined and 1 was specimen confined. Thus, low p27Klpl in poorly

differentiated tumors suggests a low likelihood of organ-confined

cancer. For such individuals, radical surgery is unlikely to be curative,and the increased risks associated with surgery may not be warranted.Conversely, higher p27Klpl staining (scores of 2-4) may better iden

tify the subset of tumors likely to be organ confined. In this study,p27K'p' staining was performed on cancers after surgical resection.Evaluation of p27Klpl staining of preoperative needle biopsy material

for presurgical prognostication of organ confinement should be undertaken in a larger study.

In lymph node-negative tumors, low p27Klpl was an independentmarker for shorter time to failure. p27K'pl staining may assist in

stratifying patients for adjuvant treatment after radical prostatectomy.Adjuvant external-beam radiotherapy lowers local failure rates after

radical prostatectomy; however, indications for this additional treatment are not always clear (39, 40). Low p27Klpl may prove to have

additional predictive value with regard to radiotherapy when it isconsidered together with other known risk factors for local recurrenceafter radical prostatectomy.

p27Klpl staining has been evaluated in previous studies by visual

scoring of the nuclear staining of tumor cells (21-23, 41). Our imageanalysis provided a second method of quantitating p27Klpl nuclear

staining. The image analysis, although considerably more time consuming, validates the visual scoring method, which can be morereadily applied to routine histological material.

The abundance of p27Klpl is important in the regulation of cyclin/cdk activity. p27Klpl binds and inhibits a broad range of cyclin/cdkcomplexes. When cultured cells exit the cell cycle, p27Kipl mRNAlevels remain constant, whereas p27Klpl protein increases, suggestingthat p27K'pl is largely posttranslationally regulated (42). Important

involvement of the ubiquitin pathway has been demonstrated (16).Studies in normal human oral mucosa and human skin keratinocytesdemonstrate an inverse relationship between cell proliferation andp27Kipl protein levels (43).5 A reduction in the level of p27Kipl could

contribute to resistance to growth inhibitors, deregulation of cellproliferation, and oncogenic change (1.2. 44). Indeed, multiple organhyperplasia, including hyperplasia of the prostate, and pituitary tumors develop in p27Klpl knockout mice, indicating p27Klpl is an

important inhibitor of cell proliferation (45-47).A relationship between low p27Klpl levels and aggressive behavior

has been demonstrated recently in various malignancies (21-24).Unlike other cdk inhibitor genes, such as p/5INK4B and p76INK4A, the/?27Klpl gene is rarely mutated in human cancers (17-20). Instead, lossof p27Klp[ appears to occur through accelerated degradation by the

ubiquitin-proteasome pathway (21, 23, 48).The increase of p27K'pl protein and/or activity by inhibitory cyto-

kines (e.g., transforming growth factor-ß) suggests that this cdk

inhibitor represents an important link between extracellular regulators

s J. M. Slingerlund. unpublished observations.

and the cell cycle (49-51). Preliminary studies suggest a role forp27K'P' ¡nandrogen-regulated prostate epithelial proliferation and

differentiation. Chen et al. (52) studied the androgen-mediated effects

on normal prostate in rats. Castration resulted in atrophy of the glanddue to apoptosis and arrest of surviving cells. Readministration oftestosterone induced epithelial cell proliferation, followed by cellulardifferentiation and an increase in p27K'pl. It was postulated that theincreased p27K'pl was playing a role in slowing proliferation as the

cells underwent differentiation.Clinical trials of NHT prior to radical prostatectomy for clini

cally localized disease consistently show reduced rates of positivesurgical margins in the treated study arms. With the immaturity offollow-up to date, none of these studies has shown a significantimprovement in disease-free survival (53-56). In our retrospective

analysis, there was a trend toward increased time to treatmentfailure following NHT. Although our numbers were small, tumorsfrom patients who received preoperative NHT expressed higherlevels of p27Klpl protein than did those from untreated patients. As

reported by others (57-59), NHT appeared to cause a shift towarda higher Gleason grade. However, the "high grade" of such tumors

is likely an artifact resulting from the distortion of tumor histologyfollowing androgen ablation. NHT has been shown to reduce theproliferative activity of tumors (58). The trend we observed towardincreased p27Klpl staining following NHT is in keeping with the

reduction in proliferation observed by others. Of note, patientswhose tumor p27Klpl remained low after NHT had a shorterrelapse-free survival (P = 0.015). Thus, low p27Kipl levels after

NHT may identify de novo hormone-resistant tumors or tumors

unlikely to have a durable response to androgen ablation. Furthermore, our data raise the provocative possibility that androgenablation therapy may be affecting the cell cycle through inhibitionof p27Klpl degradation pathways, leading to increased p27Klpl-

mediated inhibition of cyclin/cdks. Effects of hormone withdrawalon the cell cycle may precede the induction of apoptosis (60-63)

or be entirely independent of it. Unraveling the pathway forantiandrogen mediated effects on prostate cell cycle kinetics maylead ultimately to the discovery of new treatments for prostatecancer and, in particular, for hormone-refractory advanced prostate

cancer.In conclusion, we found that low p27Klpl protein in lymph node-

negative human prostate cancer is of independent prognostic value fortreatment failure after radical prostatectomy. Although the risk ratiowas 2.08, the confidence intervals were broad, reflecting the smallsample size in this study. The validity of our observations is supportedby a similar and independent observation that low p27Klpl staining

predicts poor outcome for patients with organ-confined prostate cancer made by another group (64). Among node-negative patients,p27Klpl was of even greater predictive value in the subset who

received NHT. When NHT patients were excluded from the analysis,p27KlP' was not an independent prognostic factor in the multi variate

model. However, with the small number in this subgroup, our studymay have lacked the power necessary to detect a significant result.Further studies of the prognostic value of p27Klpl in preoperative

biopsy specimens may ultimately strengthen the clinical decision thataggressive local therapy, such as radical prostatectomy, should beassigned. Further investigation may establish p27Klpl as an accurate

predictor of pathological stage prior to treatment or as a marker of theneed for adjuvant treatment postoperatively. Larger studies comparingp27K'p' levels in preoperative needle biopsies and in surgically re

sected prostate cancers following NHT may allow early detection oftumors that are refractory to androgen ablation therapy. Finally,p27Klpl and the cellular machinery that regulates its protein levels

546

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27K1PI: A PROGNOSTIC FACTOR IN PROSTATE CANCER

may represent a novel target for therapeutic intervention in prostate 27cancer.

28.

ACKNOWLEDGMENTS

We are grateful to M. Viscardi for valuable help in preparation of themanuscript.

REFERENCES

1. Parker, S. L., Tong, T., Bolden, S.. and Wingo, P. A. Cancer statistics, 1996. CACancer J. Clin., 46: 5-27, 1996.

2. National Cancer Institute of Canada. Canadian Cancer Statistics, p. 15. Toronto,Canada: National Cancer Institute of Canada. 1997.

3. Albertsen. P. C., Fryback, D. G., Storer, B. E., Kolon, T. F., and Fine, J. Long-termsurvival among men with conservatively treated localized prostate cancer. J. Am.Med. Assoc., 274: 626-631, 1995.

4. Chodak, G. W., Thisted, R. A., Gerber, G. S., Johansson, J., Adolfsson, J., Jones,G. W., Chisholm, G. D., Moskovitz, B., Livne, P. M., and Warner. J. Results ofconservative management of clinically localized prostate cancer. N. Engl. J. Med.,330: 242-248, 1994.

5. Goto, Y., Ohori, M., Arakawa, A., Rattan, M. W., Wheeler, T. M., and Scardino. P. T.Distinguishing clinically important from unimportant prostate cancers before treatment: value of systematic biopsies. J. Urol., 156: 1059-1063, 1996.

6. Partin, A. W., Kattan, M. W., Subong. E. N., Walsh, P. C., Wojno, K. J..Oesterling, J. E., Scardino, P. T., and Pearson, J. D. Combination of prostate-specific antigen, clinical stage, and Gleason score to predict pathological stage oflocalized prostate cancer. A multi-institutional update. J. Am. Med. Assoc., 277:1445-1451, 1997.

7. Hunter, T., and Pines, J. Cyclins and cancer. II. Cyclin D and cdk inhibitors come ofage. Cell, 79: 573-582, 1994.

8. Morgan, D. O. Principles of Cdk regulation. Nature (Lond.), 374: 131-134, 1995.9. Sherr, C. J. G, phase progression: cycling on cue. Cell, 79: 551-555. 1994.

10. Sherr, C. J., and Roberts, J. M. Inhibitors of mammalian G, cyclin-dependent kinases.Genes Dev., 9: 1149-1163, 1995.

11. Norbury, C., and Nurse, P. Animal cell cycles and their control. Annu. Rev. Bio-chem., 61: 441-470, 1992.

12. Solomon, M. J. Activation of the various cyclin/cdc2 proteins. Curr. Opin. Cell Biol.,5; 180-186, 1993.

13. Hartwell, L. Defects in a cell cycle checkpoint may be responsible for the genomicinstability of cancer cells. Cell, 71: 543-546, 1992.

14. Murray. A. W. Creative blocks: cell cycle checkpoints and feedback controls. Nature(Lond.). 359: 599-604, 1992.

15. Reed, S. I., Bailly. E., Dulie, V., Hengst, L., Resnitzky, D., and Slingerland, J. G,control in mammalian cells. J. Cell Sci. Suppl., 18: 69-73, 1994.

16. Pagano, M.. Tarn, S. W., Theodoras, A. M., Beer-Romero, P., Del Sal, G., Chau, V.,Yew, P. R., Draetta, G. F.. and Rolfe. M. Role of ubiquitin-proteasome pathwayin regulating abundance of the cyclin-dependent kinase inhibitor p27. Science(Washington DC), 269: 682-685, 1995.

17. Kawamata, N., Morosetti, R., Miller, C. W., Park, D., Spirin, K. S., Nakamaki, T.,Takeuchi, S., Hatta, Y., Simpson, J.. Wilczynski, S., Young, Y. L., Bartram, C. R.,and Koeffler, H. P. Molecular analysis of the cyclin-dependent kinase inhibitor genep27Kipl in human malignancies. Cancer Res.. 55: 2266-2269. 1995.

18. Ponce-Castaneda, M. V.. Lee, M-H., Latres, E., Polyak, K., Lacombe, L.,Montgomery, K., Mathew, S., Krauter, K.. Sheinfeld. J., Massague. J., and Cordon-Cardo. C. p27Klfl: chromosomal mapping to I2pl2-I2pl3.1 and absence of muta

tions in human tumours. Cancer Res., 55: 1211-1214, 1995.19. Pietenpol, J. A., Bohlander, S. K., Sato, Y., Papadopoulos, N., Liu, B., Friedman. C.,

Trask, B. J., Roberts, J. M., Kinzler, K. W., Rowley. J. D., and Vogelstein. B.Assignment of human />27Klpl gene to 12pl3 and its analysis in leukemias. Cancer

Res., 55: 1206-1210, 1995.20. Ferrando. A. A., Balbin. M.. Pendas. A. M.. Vizoso, F., Velasco, G., and Lopez-Otin,

C. Mutational analysis of the human cyclin-dependent kinase inhibitor p27kipl inprimary breast carcinomas. Hum. Genet., 97: 91-94, 1996.

21. Catzavelos, C., Bhattacharya. N., Ung, Y. C., Wilson, J. A., Roncan, L., Sandhu, C.,Shaw, P., Yeger, H., Morava-Protzner, I.. Kapusta, L., Franssen. E.. Pritchard, K. I.,

and Slingerland, J. M. Decreased levels of the cell cycle inhibitor p27Kipl protein:prognostic implications in primary breast cancer. Nat. Med.. 3: 227-230, 1997.

22. Tan, P., Cady, B., Wanner, M., Worland. P., Cukor, B., Magi-Galluzzi, C., Lavin, P.,

Draetta, G., Pagano, M., and Loda, M. The cell cycle inhibitor p27 is an independentprognostic marker in small (Tla,b) invasive breast carcinomas. Cancer Res., 57:1259-1263, 1997.

23. Loda, M.. Cukor, B., Tarn, S. W., Lavin, P.. Fiorentino, M.. Draetta, G. F., Jessup.J. M., and Pagano, M. Increased proteasome-dependent degradation of the cyclin-dependent kinase inhibitor p27 in aggressive colorectal carcinomas. Nat. Med., 3:231-234, 1997.

24. Mori, M., Mimori, K., Shiraishi, T., Tanaka, S., Ueo, H., Sugimachi, K., andAkiyoshi, T. p27 expression and gastric carcinoma. Nat. Med.. 3: 593, 1997.

25. Schroder, F. H., Hermanek, P., Denis, L., Fair, W. R., Gospodarowicz, M. K., andPavone-Macaluso, M. The TNM classification of prostate cancer. Prostate Suppl.. 4:129-138, 1992.

26. Toyoshima, H., and Hunter, T. p27, a novel inhibitor of G, cyclin-cdk protein kinaseactivity, is related to p21. Cell, 78: 67-74, 1994.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

Bacus, J. W.. and Grace, L. J. Optical microscope system for standardized cellmeasurements and analysis. Appi. Optics, 26: 3280-3293, 1987.Peto, R., Pike, M. C., Armitage, P., Breslow, N. E., Cox, D. R., Howard. S. V..Mantel. N.. McPherson. K., Peto. J., and Smith, P. G. Design and analysis ofrandomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br. J. Cancer. 35: 1-39. 1977.Cox, D. R. Regression models and life tables. J. R. Slat. Soc.. B34: 187-220. 1972.Kaplan. E. L.. and Meier. P. Nonparametric estimation from incomplete observations.Am. Stat. Assoc. J., 53: 457-481. 1958.

Oesterling, J. E., Brendler, C. B.. Epstein, J. I.. Kimball, A. W., Jr.. and Walsh, P. C.Correlation of clinical stage, serum prostatic acid phosphatase and preoperativeGleason grade with final pathological stage in 275 patients with clinically localizedadenocarcinoma of the prostate. J. Urol.. 138: 92-98, 1987.McNeal. J. E.. Villers, A. A., Redwine, E. A., Freiha, F. S., and Stamey, T. A.Histologie differentiation, cancer volume, and pelvic lymph node metastasis inadenocarcinoma of the prostate. Cancer (Phila.), 66: 1225-1233, 1990.Paulson, D. F. Impact of radical prostatectomy in the management of clinicallylocalized disease. J. Urol. /52: 1826-1830, 1994.Walsh. P. C.. Partin. A. W., and Epstein, J. I. Cancer control and quality of lifefollowing anatomical radical relropubic prostatectomy: results at 10 years. J. Urol.,152: 1831-1836. 1994.

Zincke, H., Oesterling, J. E.. Blute. M. L.. Bergslralh, E. J., Myers. R. P., and Barrett,D. M. Long-term (15 years) results after radical prostatectomy for clinically localized(stage T2c or lower) prostate cancer. J. Urol., 752: 1850-1857, 1994.Murphy. G. P., Mettlin. C.. Menck. H.. Winchester, D. P.. and Davidson. A. M.National patterns of prostate cancer treatment by radical prostatectomy: results of asurvey by the American College of Surgeons Commission on Cancer. J. Urol., /52:1817-1819, 1994.Gerber, G. S.. Thisted, R. A., Scardino, P. T., Frohmuller. H. G. W.. Schroeder,F. H., Paulson, D. F., Middlelon, A. W.. Jr., Rukstalis, D. B., Smith. J. A., Jr.,Schellhammer, P. F., Ohori. M., and Chodak. G. W. Results of radical prostatectomyin men with clinically localized prostate cancer. J. Am. Med. Assoc., 276: 615-619,1996.Ohori. M.. Goad. J. R., Wheeler. T. M., Eastham. J. A.. Thompson, T. C., andScardino, P. T. Can radical prostatectomy alter the progression of poorly differentiated prostate cancer? J. Urol.. /52: 1843-1849, 1994.Coetzee, L. J., Hars, V.. and Paulson. D. F. Postoperative prostate-specific antigenas a prognostic indicator in patients with margin-positive prostate cancer, undergoing adjuvant radiotherapy after radical prostatectomy. Urology. 47: 232-235,1996.Schild, S. E.. Wong. W. W., Grado. G. L.. Buskirk, S. J.. Robinow, J. S., Frick,L. M.. and Ferrigni. R. G. Radiotherapy for isolated increases in serum prostate-specific antigen levels after radical prostatectomy. Mayo Clin. Proc.. 69: 613-619, 1994.Porter, P. L.. Malone. K. E.. Heagerty, P. J.. Alexander, G. M., Gatti, L. A., Firpo.E. J.. Dating, J. R.. and Roberts. J. M. Expression of cell cycle regulators p27kipl andcyclin E, alone and in combination, correlate with survival in young breast cancerpatients. Nat. Med.. 3: 222-225, 1997.

Hengst, L., and Reed, S. I. Translational control of p27Kipl accumulation during thecell cycle. Science (Washington DC), 27/: 1861-1864, 1996.

Jordan. R.. Bradley. G.. and Slingerland, J. M. Reduced levels of (he cell cycleinhibitor p27k'pl in epithelial dysplasia and carcinoma of the oral cavity. Am. J.

Pathol., in press. 1998.Bostwick. D. G. Neoplasms of the prostate. In: D. G. Bostwick and J. N. Eble (eds.).Urologie Surgical Pathology, p. 343-421. St. Louis, MO: Mosby, 1997.

Nakayama, K., Ishida, N., Shirane, M.. Inomala. A.. Inoue. T., Shishido. N., Horii, I.,Loh, D. Y.. and Nakayama. K. Mice lacking p27Kipl display increased body size,multiple organ hyperplasia, retinal dysplasia. and pituitary tumors. Cell. 85: 707-720,1996.Fero, M. L., Rivkin. M.. Tasch. M., Porter, P.. Carow, C. E., Polyak. K.. Firpo. E..Tsai, L., Broudy, V., Perlmutter. R. M., Kaushansky. K.. and Roberts, J. M. Asyndrome of multi-organ hyperplasia with features of gigantism. tumorigenesis andfemal sterility in p27Kipl-deficient mice. Cell, 85: 733-744, 1996.Kiyokawa. H.. Kineman. R. D.. Manova-Todorova, K. O.. Scares, V. C.. Hoffman.E. S., Ono. M.. Mi.ni.un. D., Hayday, A. C.. Frohman. L. A., and Koff, A. Enhancedgrowth of mice lacking the cyclin-dependent kinase inhibitor function of p27Kipl.Cell, 85: 721-732, 1996.

Esposito. V.. Baldi, A., De Luca, A., Groger, A. M., Loda, M.. Giordano, G. G.,Caputi, M.. Baldi. F.. Pagano, M., and Giordano, A. Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res., 57:3381-3385, 1997.Slingerland, J. M.. Hengst, L., Pan, C-H., Alexander. D., Stampfer, M. R., and Reed,S. I. A novel inhibitor of cyclin-Cdk activity detected in transforming growth factorß-arrestedepithelial cells. Mol. Cell. Biol., 14: 3683-3694. 1994.

Polyak. K., Kalo. J. Y.. Solomon. M. J., Sherr. C. J., Massague. J.. Roberts, J. M., andKoff, A. p27Kipl, a cyclin-Cdk inhibitor, links transforming growth factor-ßandcontact inhibition to cell cycle arrest. Genes Dev., 8: 9-22, 1994.Koff, A.. Ohtsuki. M., Polyak, K., Roberts. J. M., and Massague. J. Negativeregulation of G, in mammalian cells: inhibition of cyclin E-dependenl kinase byTGF-ß.Science (Washington DC), 260: 536-539, 1993.Chen, Y., Robles, A. I., Martinez, L. A., Liu. F.. Gimenez-Conti. I. B.. and Conti, C. J.Expression of G, cyclins, cyclin-dependent kinases. and cyclin-dependent kinaseinhibitors in androgen-induced prostate proliferation in castrated rats. Cell GrowthDiffer.. 7: 1571-1578, 1996.

Labrie, F., Dupont, A.. Cusan. L., Gomez, J.. Diamond. P.. Koutsilieris, M.. Suburu.R., Fradet. Y.. Lemay, M.. Tetu. B.. Emond, J.. and Candas. B. Downstaging of

547

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

LOSS OF P27KIP1: A PROGNOSTIC FACTOR IN PROSTATE CANCER

localised prostate cancer by neoadjuvanl therapy with flutamide and lupron: the firstcontrolled and randomized trial. Clin. Invest. Med., 16: 499-509, 1993.

54. Van Poppel, H.. De Ridder, D., Elgamal, A. A.. Van de Voorde. W.. Werbrouck, P..Ackaert, K., Oyen. R., Pittomvils, G., and Baert. L. Neoadjuvant hormonal therapybefore radical prostatectomy decreases the number of positive surgical margins instage T2 prostate cancer: interim results of a prospective randomized trial. TheBelgian Uro-Oncological Study Group. J. Urol., 154: 429-434, 1995.

55. Soloway, M. S., Sharifi, R.. Wajsman. Z., McLeod, D., Wood. D. P.. Jr.. andPuras-Baez. A. Randomized prospective study comparing radical prostatectomy alone

versus radical prostatectomy preceded by androgen blockade in clinical stage B2(T2bNxMO) prosiate cancer. The Lupron Depot Neoadjuvant Prostate Cancer StudyGroup. J. Urol.. 154: 424-428. 1995.

56. Goldenberg, S. L., Klotz. L. H.. Srigley, J.. Jewett, M. A., Mador, D.. Fradet, Y..Barkin, J., Chin, J.. Paquin, J.. Bullock. M. J., and Laplanle, S. Randomized,prospective, controlled study comparing radical prostatectomy alone and neoadjuvanlandrogen withdrawal in the treatment of localized prostate cancer. Canadian UrologieOncology Group. J. Urol.. 156: 873-877, 1996.

57. Murphy, W. M., Soloway, M. S., and Barrows, G. H. Pathologic changes associatedwith undrogen deprivation therapy for prostate cancer. Cancer (Phila.), 68: 821-828,1991.

58. Armas, O. A.. Aprikian. A. G.. Melamed, J., Cordon-Cardo. C, Cohen. D. W.,Erlandson, R., Fair, W., R., and Reuter, V. E. Clinical and pathobiological effects of

neoadjuvant total androgen ablation therapy on clinically localized prostatic adeno-carcinoma. Am. J. Surg. Pathol., 18: 979-991, 1994.

59. Civantos, F., Martial. M. A., Banks. E. R., Ho, C.. Speights, V. O., Drew. P. A.,Murphy, W. M., and Soloway, M. Pathology of androgen deprivation therapy inprostate carcinoma: a comparative study of 173 patients. Cancer (Phila.), 75: 1634-

1641. 1995.60. Kyprianou, N., English, H. F., and Isaacs, J. T. Programmed cell death during

regression of PC-82 human prostate cancer following androgen ablation. Cancer Res..50. 3748-3753, 1990.

61. Kyprianou, N.. and Isaacs, J. T. Activation of programmed cell death in the rat ventralprostate after castration. Endocrinology. 122: 552-562, 1988.

62. English. H. F.. Kyprianou. N., and Isaacs. J. T. Relationship between DNA fragmentation and apoptosis in the programmed cell death in the rat prostate followingcastration. Prostate. 15: 233-250. 1989.

63. Kyprianou. N., and Isaacs, J. T. Expression of transforming growth factor-ßin the ratventral prostate during castration-induced programmed cell death. Mol. Endocrinol..3: 1515-1522, 1989.

64. Yang. R. M., Nailoh, J., Philipson, J., Wang. H.. deKemion, J. B.. Loda, M., andReiter, R. E. Low levels of p27 protein expression predict poor disease-free survivalin patients with pathological T2a-T3b adenocarcinoma of the prostate. J. Urol., in

press, 1998.

54S

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1998;58:542-548. Cancer Res   John Tsihlias, Linda R. Kapusta, Gerrit DeBoer, et al.   AdenocarcinomaPrognostic Factor in Localized Human Prostate

Is a NovelKip1Loss of Cyclin-dependent Kinase Inhibitor p27

  Updated version

  http://cancerres.aacrjournals.org/content/58/3/542

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/3/542To request permission to re-use all or part of this article, use this link

on April 14, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from