Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
[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 Science Center: Department of Politologi: Women's College Hospital ¡I.Z.]; and Departments of Medical Oncology ¡J.M. SJ and Clinical Trials and Epidemiology ¡G.D./. TorontoSunnvbrook 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 was undertaken to assess the prognostic value of p27Klpl in localized human prostate cancer. Archival material from 113 radical prostatectomy spec imens obtained between 1985 and 1993 was stained immunohistochemically 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 biochem ical and clinical recurrence, and survival. Strong p27Klpl staining was uniformly seen in benign prostatic epithelial components in all tumor sections. p27Kipl staining was reduced in most prostate cancers and was variable in prostatic intraepithelial neoplasia. Decreased p27Kipl staining «25'r of nuclei stained positive for p27KIpl) correlated with seminal vesicle involvement (/' = 0.0032) and with higher Gleason grade I/' = 0.0114). On univariate analysis, low p27Kipl predicted an increased risk of treatment failure in the node-negative cohort (/' = 0.0037) and in the 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 follow ing radical retropubic prostatectomy (n = 102; /" = 0.047). Seminal vesicle involvement (/' = 0.034) and positive surgical margins (/' = 0.047) were also independent prognostic factors for disease recurrence. In patients who received preoperative neoadjuvant hormonal therapy, low p27Klpl in the pathological specimen was an even stronger predictor of outcome than it was in the entire group (n = 23, P = 0.015). INTRODUCTION Prostate carcinoma is the most common solid malignancy and the second most common cause of cancer-related death in men. There were 337,000 estimated new cases of prostate cancer and 45,400 estimated 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 and aggressive (3-5). Traditional parameters used to predict extent of disease and natural history include: clinical stage, tumor grade, and serum PSA.4 Published nomograms allow more accurate prediction of based on these factors is difficult. A more accurate prognostic marker for this patient group would permit assignment of more aggressive treatment to those most likely to fail and would justify expectant management (deferral of active treatment) in those likely to have indolent disease. There is increasing evidence that cell cycle regulators are disrupted in human cancers (7). The cell cycle is governed by cdks, the activities of which are regulated by binding of positive effectors, the cyclins (8, 9); by negative regulators, the cdk inhibitors (10); and by phosphorylation and dephosphorylation events (11, 12). The cdks integrate mitogenic and growth-inhibitory signals and coordinate cell cycle transitions (13, 14). Passage through G, into S phase is regulated by the activities of cyclin D-, cyclin E-, and cyclin A-associated kinases. B-type cyclin-associated kinases regulate G2-M. Two families of cdk inhibitors 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, in cluding p21cipl, p27Kipl, and p57Kip2, bind and inhibit their targets. The cdk inhibitors regulate entry into and progression through the cell cycle by modulating the activity of cdks in response to mitogenic and antimitogenic stimuli. The cyclins, cdks, and cdk inhibitors are fre quently altered in cancer or disrupted secondarily by other oncogenic events (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 at the 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 in human tumors (17-20), decreased p27Klpl protein has potential im portance as a prognostic factor in breast (21, 22), colon (23), and gastric carcinoma (24), and its reduction may be associated with tumor progression. Here, we demonstrate that the loss of p27Klpl appears to be of prognostic carcinoma. value in primary resectable prostate pathological stage using these three parameters (6). Most patients, however, fall in a middle zone where meaningful prognostication PATIENTS 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 with 18 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 Fox Run, and Cancer Care Ontario. 2 The first -1Requests National Cancer Institute of Canada, supported by funding from the Terry was also supported by the Toronto-Sunnybrook Regional Cancer Center and Ontario. J. M. S. is a Clinician Investigator, supported by Cancer Care two authors contributed equally to this work. for reprints should be addressed to either Dr. L. Kapusta. Department of Pathology, or Dr. J. Slingerland. Cancer Biology Research, Sunnybrook Health Science Center, 2075 Bayview Avenue, Toronto, Ontario, Canada, M4N 3M5. Phone: (416) 4804120 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 Analysis Systems; NHT, neoadjuvant hormonal therapy; KIP, kinase inhibitor protein. AND METHODS Patient Population. The study group consisted of 113 patients who had undergone radical prostatectomy, performed by one of two surgeons, at Sun nybrook Health Science Center between January 1986 and April 1993 and for whom clinical follow-up data were available. Paraffin blocks from the embed ded radical prostatectomy specimens were selected on the basis of having both benign and malignant histology present within the same tumor section. The clinical database prepared from patient records included preoperative PSA, clinical and pathological staging, Gleason tumor grade, time to treatment failure, and survival. The patients were staged according to the TNM system (25). Treatment failure was defined as any one of the following: institution of hormonal therapy postoperatively. failure of the PSA to return to undetectable levels postoperatively (<0.2 /xg/liter; Hybritech assay), or any new PSA elevation above undetectable levels on two successive readings during post operative follow-up. Patients who had lymph node metastasis at the time of 542 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27IÃœPI:A PROGNOSTIC FACTOR surgery were treated routinely with postoperative (adjuvant) hormonal therapy. Because institution of hormone therapy is within our definition of treatment failure, node-positive eases were excluded from the prognostic factor analysis. Immunohistochemistry. Expression of p27Klpl protein was determined IN PROSTATE CANCER Table 1 Clinical and pathnloaical Parametern(%)Preoperative (80.5)163420796501794816422694 factorsPreoperative PSA*<44-1010-20>20Clinical immunohistochemically. Paraffin sections of tumor blocks were deparaffinized with xylene. rehydrated, and microwaved for 10 min in 10 mM citrate buffer (pH 6.0). Sections were blocked with 3% hydrogen peroxide in methanol followed by normal horse serum (1:20 dilution) in 0.1 M PBS at pH 6.0 and then incubated overnight at 4°Cwith anti-p27Kipl monoclonal antibody (Trans- (27)12(12)11 T-stage'TibTicT2aT2bT2cNeedle (10)9(8)56 duction Laboratories, Lexington, KY) diluted 1:1000 (0.25 ng/ml) in PBS. Slides were then reacted with biotin-labeled anti-mouse IgG and incubated with preformed avidin-biotin-peroxidase complex (Vector Laboratories, Burlingame. CA). Metal-enhanced diaminobenzidine substrate (Pierce, Rock- (50)25 (22)12(11)57(66)22(25)8(9)24(21)89 grade2-678-10Neoadjuvant biopsy Gleason ford, IL) was then added in the presence of horseradish peroxidase. Sections were counterstained with hematoxylin, dehydrated, and mounted. The degree of 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 therapyYesNoPostoperative hormonal (79)49 benign and malignant tissue. Negative controls using isotype-specific polyclonal mouse immunoglobulin showed no staining. Furthermore, the pattern of p27 staining in tumor samples obtained with the monoclonal antibody from Transduction Laboratories was confirmed using three different polyclonal p27Klpl sera (one from Santa Cruz Biotechnology, Santa Cruz, CA; and two factorsGleason tumor2-678-10Extra grade from resected (43)45 (40)19(17)76 extensionYesNoMargins capsular kindly provided by T. Hunter, the Salk Institute. La Jolla. CA; Ref. 26). p27 staining with the Santa Cruz antibody was lost when the antibody was preadsorbed with blocking p27 peptide (Ref. 21 and data not shown). Between 15 and 20 high-power fields of tumor were scored for the percentage of nuclei showing positive p27Klpl staining. The scoring system used was as in our previous studies (21): 1, 0-25%; 2, 25-50%; 3, 50-75%; data p27"22(19.5)26652368396422069718418410123192 (%)113(100)18(19)40(42)26 p2791 (67)37 (33)70 positiveYesNoSeminal (62)43 (38)21 involvementYesNoLymph vesicle and 4, >75% (19)92(81)11(10)102 positive nuclear staining. Image Analysis. Image cytometric analysis was performed with the CAS 200 (Becton Dickinson Cellular Imaging Systems, San Jose, CA), a videobased interactive image cytometer (27). using the Quantitative Nuclear Anti gen Program software (Version 3.06; CAS Inc., Becton Dickinson Cellular Imaging Systems). Four negative fields on each slide were analyzed to estab lish the nuclear and antibody threshold optical densities by which the image analyzer could discriminate positive nuclear staining from background. To provide a second independent method to verify nuclear p27Klpl staining scores, involvementYesNoCurrent node (90)104(92)3(3)6(5)<25% statusAliveDead unknown)Dead (cancer or cause (other cause)Total 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). image analysis was performed on selected cases scored previously by percent age 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 were selected 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 minimum total 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) and Cox'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 a function of time after diagnosis. Correlation analysis between prognostic factors was done using Spearman and Pearson coefficients. tered preoperatively to 24 of 113 patients in a variable manner. The drug regimens used were cyproterone acetate (14 patients), diethylstilbesterol (5 patients), leuprolide plus nilutamide (2 patients), finasteride (2 patients), and flutamide alone ( I patient). Duration of hor monal 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 of radical prostatectomy (33-35). Immunohistochemistry Results and Image Analysis. Strong ex pression of p27K'pl was consistently noted in virtually all nuclei of benign prostatic epithelium, and there was a consistent but variable degree of reduction in both the intensity of p27K'pl staining and the number of nuclei stained in prostate carcinoma (Fig. 1). Although the nuclei of prostatic intraepithelial neoplasia tended to be p27Klpl RESULTS Clinical and Pathological Cohort Data. Median patient age was 65 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, scores 2-6, score of 7, and scores 8-10, based on the observation that a Gleason score of 7 confers a significantly worse prognosis than does a score of 6 (31, 32). The preoperative Gleason score was known for 87 patients; preoperative serum PSA was available in 96 patients. All other data were available for the entire cohort (n = 113). The majority of patients were clinical stage T2 (n = 93), and the remainder were T, (n = 20). Neoadjuvant hormonal therapy (NHT) had been adminis- 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 immunohistochemistry. Image highest values of range, 144-186; each category of analysis of benign nuclei consistently showed the summated absorbance in all cases studied (n = 12; mean, 154.4). Over 150 nuclei were evaluated in 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 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27Klpi: A PROGNOSTIC FACTOR IN PROSTATE CANCER K**»**^ Fig. I. p27K'pl nuclear staining of prostale adenocarcinoma. Paraffin-embedded material from radical prostatectomy specimens was stained for p27KlP' and counterstained with hematoxylin. The ~c- percentage of malignant nuclei with positive nu clear staining was scored independently by two pathologists into four categories. Large arrow heads, benign prostatic glandular epithelium; small arrows, prostatic adenocarcinoma. A. adenocarcinoma of prostate, Gleason score 4. with no nuclear staining 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, adenocarci noma of prostate. Gleason score 3. with p27Klpl nuclear staining category 3 (50-75% nuclei posi tive). D, adenocarcinoma of prostate, Gleason score 3 with high nuclear p27Klpl staining, category 4 (>75% nuclear staining). Magnification (A-D), X400. %*A •¿*' A » Table 2 Correlation of p27K'1'' with other prognostic parameters" p27K"" n Univariate analysis of lymph node-negative cases (n = 102) by Cox regression 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 positive surgical margins (P = 0.030; Table 3A). Correlation coefficient All113)Seminal patients (n = involvementGleason vesicle tumorMargins grade from resected positivePreoperative coefficient)Extracapsular PSA ' (Pearson extensionNHTLymph involvementAge node coefficient)Preoperative, (Pearson Table 3 Univariate analysis for time to treatment failure Prognostic marker 89)Seminal NHT excluded (n = involvementGleason vesicle tumorMargins grade from resected positivePreoperative coefficient)Eixtracapsular PSA* (Pearson P" Node-negative cases (n = 102)* 140.0930.1850.1440.550.51 Seminal vesicle involvement 0.0030 <25% nuclei p27 positive 0.0037 Gleason grade from resected tumor (2-6, 7, 8-10) 0.0094 extensionLymph Margins positive 0.030 involvementAge node Needle biopsy gleason grade (2-6. 7. 8-10; n = 78) 0.051 (Pearson coefficient)113113113%11311311311389898973898989-0.340-0.202-0.201-0.198-0.153+0.146-0.065-0.055-0.309-0.267-0.179-0.157-0.156-0.064-0.0710.00020.0320.0330.0540.1070.1230.500.560.00 Neoadjuvant hormonal therapy 0.25 " Low p27Klpl, <25% nuclear staining; high p27Klpl, s25% nuclear staininig; SpearClinical T-stager 0.26rf Preoperative PSA' (<4, 4-10, 10-20, >20; n = 86) 0.35d man rank correlation coefficient, except where otherwise noted. * In fig/liter. Extracapsular extension 0.38 Age 0.42 Node-negative cases, excluding NHT (n = 79) Gleason grade from resected tumor (2-6, 7, 8-10) Needle biopsy Gleason grade (2-6. 7. 8-10; n = 58) Seminal vesicle involvement <25% nuclei p27 positive Margins positive Extracapsular extension Preoperative PSA' (<4, 4-10, 10-20, >20; n = 64) Clinical T-stage' (25-50% nuclear staining) or 3 (50-75% nuclear staining), had summated absorbance values between those of groups 1 and 4: cases scored as 2 had a range of 58-71 and a mean of 63.0, and cases scored as 3 had a range of 83-92 and a mean of 88.3. Prognostic Role of p27Kipl. In a correlation analysis between prognostic factors, low 27Klpl was strongly associated with seminal vesicle involvement and with increased Gleason grade (P = 0.0002 and P = 0.032 respectively; Table 2). A trend toward increasing PSA was observed when p27Kipl was low (P = 0.054). Low p27Kipl did Age " 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. not correlate significantly with extracapsular extension, lymph node involvement, use of NHT, or patient age (Table 2). 0.0027 0.012 0.044 0.049 0.059 0.30 O^ 0.65d 0.74 ' In 544 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27KIHI: A PROGNOSTIC FACTOR IN PROSTATE CANChK only those tumors showing low p27Klpl nuclear staining (<25% nuclei positive) had a distinctly worse prognosis over the follow-up period than did the remainder of the cohort (P = 0.0037; Fig. 25, Kaplan-Meier curves). The probabilities of being free of treatment failure at 5 years were 38 ± 23% for low-p27Klpl tumors and 72 ±10% for the high-p27Kipl tumors. Multivariate analysis of node-negative tumors (n = 102) identified low p27Klpl as a novel independent prognostic factor for time to treatment failure (P = 0.047; risk ratio = 2.08; Table 4). Seminal 01 9 234567 10 11 vesicle involvement and positive surgical margins were also of inde pendent prognostic value, with risk ratios of 2.29 (P = 0.034) and 2.07 (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 did not reach statistical significance (P = 0.25; Table 3A). A trend toward increased p27Klpl staining was observed in tumors from patients who 12 received NHT. Only 2 of 24 tumors (8.3%) resected following androgen ablation therapy had low p27Klpl (<25% positive nuclei), 100 : whereas 20 of 89 tumors (22.5%) that did not receive NHT had low p2yKipi yjjjj, (Jifferengg was nO( statistically significant (P = 0.15; two-tailed Fisher's exact test). NHT had a significant effect on the 80 distribution of grade in the resected specimens. Among the 89 patients who 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 of 2-6, 11 (46%) had Gleason scores of 7, and 8 (33%) had Gleason scores of 8-10. This trend toward increased Gleason grade following NHT was statistically significant ()f for trend = 8.7; P = 0.003). Interestingly, when node-negative cancers treated with preoperative NHT were analyzed separately, low p27Klp' in the resected tumor was 60 40 : 20 25% p27 p = 0.0037 01 23456789 10 11 an even more significant predictor of reduced time to treatment failure than it was in the group as a whole (P = 0.015; n = 23). When the patients who received NHT were removed from the analysis, low p27Kipl still predicted reduced time to relapse (P = 0.049; Table 3B and Fig. 2C, Kaplan-Meier curve). On multivariate analysis of the cases with negative lymph nodes who did not receive NHT (n = 79), only Gleason grade proved to be a statistically significant prognostic factor for relapse-free rate (/> = 0.0034; risk ratio = 2.10). 12 Years 100 - 80 ; DISCUSSION >25%p271 '-m— «---h 6J0'S0)i1 +H ¡ Established modalities for managing localized prostate cancer in clude radical prostatectomy, radical radiation therapy, and surveil lance. Untreated low-grade cancers in men with life expectancies of less than 10-15 years are not likely to affect longevity (3). When radical therapy is not given to patients with tumors of moderate or high grade, there is a loss in life years of 4-5 and 6-8 years, respectively (3). Radical prostatectomy has a low mortality but sig nificant morbidity, with impotence resulting in 50-70% and stress incontinence in 2-6% of cases (34-36). In tumors treated by radical H<25%p27p :g2040 ;1 01 = 0.049 23456789 10 11 12 Years Fig. 2. A, Kaplan-Meier curve of time to treatment failure (n = 102). B, Kaplan-Meier curves of time to treatment failure, comparing patients with <25% p27Klpl nuclear staining 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 received NHT, comparing <25% p27Klpl nuclear staining to those with 225% p27Klpl nuclear staining (n = 79; P = 0.049). Table 4 Multivariate analysis for lime to treatment failure Model for treatment failure Risk ratio (CD" Node-negative cases (n = 102) (1.07-4.92) Seminal vesicle involvement 2.08 (1.01-4.27) <25% p27 positive nuclei (1.01-4.24)2.10'' 2.07 Margins positive NHT recipients excluded (n = 79) (1.28-3.44)0.034 Gleason grade from resected tumor (2-6. 7. 8-10)2.29 " CI. 95% confidence interval for the relative risk ratio. h From Wald x~' Per step. The Kaplan-Meier curve of time to treatment failure for nodenegative study patients showed a 5-year actuarial failure-free rate of 66% (±9.6%, 95% confidence interval; Fig. 14). When the diseasefree rate was plotted versus p27Klpl staining scores of 1, 2, 3, or 4, 0.047 0.0470.0034 545 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27lap': A PROGNOSTIC FACTOR IN PROSTATE CANCER prostatectomy, increasing grade predicts a shorter disease-free sur vival (37). When the cancer is organ confined, however, disease-free survival is excellent, regardless of tumor grade (38). Because radical prostatectomy is unlikely to be curative in locally advanced disease, a marker 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 extracapsular 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 re maining 12 high-grade tumors without low p27Klpl, 3 were organ confined 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 and the cell cycle (49-51). Preliminary studies suggest a role for p27K'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 gland due to apoptosis and arrest of surviving cells. Readministration of testosterone induced epithelial cell proliferation, followed by cellular differentiation and an increase in p27K'pl. It was postulated that the increased 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 positive surgical margins in the treated study arms. With the immaturity of follow-up to date, none of these studies has shown a significant improvement in disease-free survival (53-56). In our retrospective analysis, there was a trend toward increased time to treatment failure following NHT. Although our numbers were small, tumors from patients who received preoperative NHT expressed higher levels of p27Klpl protein than did those from untreated patients. As reported by others (57-59), NHT appeared to cause a shift toward a higher Gleason grade. However, the "high grade" of such tumors for presurgical prognostication of organ confinement should be un dertaken in a larger study. In lymph node-negative tumors, low p27Klpl was an independent marker for shorter time to failure. p27K'pl staining may assist in is likely an artifact resulting from the distortion of tumor histology following androgen ablation. NHT has been shown to reduce the proliferative activity of tumors (58). The trend we observed toward increased p27Klpl staining following NHT is in keeping with the stratifying patients for adjuvant treatment after radical prostatectomy. Adjuvant external-beam radiotherapy lowers local failure rates after reduction in proliferation observed by others. Of note, patients whose tumor p27Klpl remained low after NHT had a shorter relapse-free survival (P = 0.015). Thus, low p27Kipl levels after radical prostatectomy; however, indications for this additional treat ment are not always clear (39, 40). Low p27Klpl may prove to have NHT may identify de novo hormone-resistant tumors or tumors additional predictive value with regard to radiotherapy when it is considered together with other known risk factors for local recurrence after radical prostatectomy. p27Klpl staining has been evaluated in previous studies by visual unlikely to have a durable response to androgen ablation. Further more, our data raise the provocative possibility that androgen ablation therapy may be affecting the cell cycle through inhibition of p27Klpl degradation pathways, leading to increased p27Klpl- scoring of the nuclear staining of tumor cells (21-23, 41). Our image analysis provided a second method of quantitating p27Klpl nuclear mediated inhibition of cyclin/cdks. Effects of hormone withdrawal on the cell cycle may precede the induction of apoptosis (60-63) staining. The image analysis, although considerably more time con suming, validates the visual scoring method, which can be more readily 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/cdk complexes. When cultured cells exit the cell cycle, p27Kipl mRNA levels remain constant, whereas p27Klpl protein increases, suggesting that p27K'pl is largely posttranslationally regulated (42). Important or be entirely independent of it. Unraveling the pathway for antiandrogen mediated effects on prostate cell cycle kinetics may lead ultimately to the discovery of new treatments for prostate cancer and, in particular, for hormone-refractory advanced prostate involvement of the ubiquitin pathway has been demonstrated (16). Studies in normal human oral mucosa and human skin keratinocytes demonstrate an inverse relationship between cell proliferation and p27Kipl protein levels (43).5 A reduction in the level of p27Kipl could contribute to resistance to growth inhibitors, deregulation of cell proliferation, and oncogenic change (1.2. 44). Indeed, multiple organ hyperplasia, including hyperplasia of the prostate, and pituitary tu mors develop in p27Klpl knockout mice, indicating p27Klpl is an cancer. In conclusion, we found that low p27Klpl protein in lymph nodenegative human prostate cancer is of independent prognostic value for treatment failure after radical prostatectomy. Although the risk ratio was 2.08, the confidence intervals were broad, reflecting the small sample size in this study. The validity of our observations is supported by a similar and independent observation that low p27Klpl staining predicts poor outcome for patients with organ-confined prostate can cer 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 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, loss of 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 cytokines (e.g., transforming growth factor-ß) suggests that this cdk inhibitor represents an important link between extracellular regulators s J. M. Slingerlund. unpublished observations. model. However, with the small number in this subgroup, our study may 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 that aggressive local therapy, such as radical prostatectomy, should be assigned. Further investigation may establish p27Klpl as an accurate predictor of pathological stage prior to treatment or as a marker of the need for adjuvant treatment postoperatively. Larger studies comparing p27K'p' levels in preoperative needle biopsies and in surgically re sected prostate cancers following NHT may allow early detection of tumors that are refractory to androgen ablation therapy. Finally, p27Klpl and the cellular machinery that regulates its protein levels 546 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27K1PI: A PROGNOSTIC may represent a novel target for therapeutic intervention in prostate cancer. IN PROSTATE CANCER 27 ACKNOWLEDGMENTS We are grateful to M. Viscardi for valuable help in preparation manuscript. FACTOR of the REFERENCES 1. Parker, S. L., Tong, T., Bolden, S.. and Wingo, P. A. Cancer statistics, 1996. CA Cancer 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-term survival 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 of conservative 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 treat ment: 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 prostatespecific antigen, clinical stage, and Gleason score to predict pathological stage of localized 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 of age. 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. Biochem., 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 genomic instability 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 pathway in 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 gene p27Kipl 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 CordonCardo. 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 in primary 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 independent prognostic 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 cyclindependent 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., and Akiyoshi, 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., and Pavone-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 kinase activity, is related to p21. Cell, 78: 67-74, 1994. Bacus, J. W.. and Grace, L. J. Optical microscope system for standardized cell measurements and analysis. Appi. Optics, 26: 3280-3293, 1987. 28. 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 of randomized clinical trials requiring prolonged observation of each patient. II. Anal ysis 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. 31. 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 preoperative Gleason grade with final pathological stage in 275 patients with clinically localized adenocarcinoma of the prostate. J. Urol.. 138: 92-98, 1987. 32. 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 in adenocarcinoma of the prostate. Cancer (Phila.), 66: 1225-1233, 1990. 33. Paulson, D. F. Impact of radical prostatectomy in the management of clinically localized disease. J. Urol. /52: 1826-1830, 1994. 34. Walsh. P. C.. Partin. A. W., and Epstein, J. I. Cancer control and quality of life following anatomical radical relropubic prostatectomy: results at 10 years. J. Urol., 152: 1831-1836. 1994. 35. 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. 36. 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 a survey by the American College of Surgeons Commission on Cancer. J. Urol., /52: 1817-1819, 1994. 37. 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 prostatectomy in men with clinically localized prostate cancer. J. Am. Med. Assoc., 276: 615-619, 1996. 38. Ohori. M.. Goad. J. R., Wheeler. T. M., Eastham. J. A.. Thompson, T. C., and Scardino, P. T. Can radical prostatectomy alter the progression of poorly differenti ated prostate cancer? J. Urol.. /52: 1843-1849, 1994. 39. Coetzee, L. J., Hars, V.. and Paulson. D. F. Postoperative prostate-specific antigen as a prognostic indicator in patients with margin-positive prostate cancer, under going adjuvant radiotherapy after radical prostatectomy. Urology. 47: 232-235, 1996. 40. 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 prostatespecific antigen levels after radical prostatectomy. Mayo Clin. Proc.. 69: 613619, 1994. 41. 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 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nat. Med.. 3: 222-225, 1997. 42. Hengst, L., and Reed, S. I. Translational control of p27Kipl accumulation during the cell cycle. Science (Washington DC), 27/: 1861-1864, 1996. 43. Jordan. R.. Bradley. G.. and Slingerland, J. M. Reduced levels of (he cell cycle inhibitor p27k'pl in epithelial dysplasia and carcinoma of the oral cavity. Am. J. Pathol., in press. 1998. 44. 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. 45. 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. 46. 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. A syndrome of multi-organ hyperplasia with features of gigantism. tumorigenesis and femal sterility in p27Kipl-deficient mice. Cell, 85: 733-744, 1996. 47. 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. Enhanced growth of mice lacking the cyclin-dependent kinase inhibitor function of p27Kipl. Cell, 85: 721-732, 1996. 48. 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 cyclindependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res., 57: 3381-3385, 1997. 49. 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 factor ß-arrestedepithelial cells. Mol. Cell. Biol., 14: 3683-3694. 1994. 50. Polyak. K., Kalo. J. Y.. Solomon. M. J., Sherr. C. J., Massague. J.. Roberts, J. M., and Koff, A. p27Kipl, a cyclin-Cdk inhibitor, links transforming growth factor-ßand contact inhibition to cell cycle arrest. Genes Dev., 8: 9-22, 1994. 51. Koff, A.. Ohtsuki. M., Polyak, K., Roberts. J. M., and Massague. J. Negative regulation of G, in mammalian cells: inhibition of cyclin E-dependenl kinase by TGF-ß.Science (Washington DC), 260: 536-539, 1993. 52. 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 kinase inhibitors in androgen-induced prostate proliferation in castrated rats. Cell Growth Differ.. 7: 1571-1578, 1996. 53. 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 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. LOSS OF P27KIP1: A PROGNOSTIC 54. 55. 56. 57. 58. FACTOR localised prostate cancer by neoadjuvanl therapy with flutamide and lupron: the first controlled and randomized trial. Clin. Invest. Med., 16: 499-509, 1993. 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 therapy before radical prostatectomy decreases the number of positive surgical margins in stage T2 prostate cancer: interim results of a prospective randomized trial. The Belgian Uro-Oncological Study Group. J. Urol., 154: 429-434, 1995. Soloway, M. S., Sharifi, R.. Wajsman. Z., McLeod, D., Wood. D. P.. Jr.. and Puras-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 Study Group. J. Urol.. 154: 424-428. 1995. 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 neoadjuvanl androgen withdrawal in the treatment of localized prostate cancer. Canadian Urologie Oncology Group. J. Urol.. 156: 873-877, 1996. Murphy, W. M., Soloway, M. S., and Barrows, G. H. Pathologic changes associated with undrogen deprivation therapy for prostate cancer. Cancer (Phila.), 68: 821-828, 1991. 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 IN PROSTATE CANCER neoadjuvant total androgen ablation therapy on clinically localized prostatic adenocarcinoma. 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 in prostate carcinoma: a comparative study of 173 patients. Cancer (Phila.), 75: 16341641. 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 ventral prostate after castration. Endocrinology. 122: 552-562, 1988. 62. English. H. F.. Kyprianou. N., and Isaacs. J. T. Relationship between DNA fragmen tation and apoptosis in the programmed cell death in the rat prostate following castration. Prostate. 15: 233-250. 1989. 63. Kyprianou. N., and Isaacs, J. T. Expression of transforming growth factor-ßin the rat ventral 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., and Reiter, R. E. Low levels of p27 protein expression predict poor disease-free survival in patients with pathological T2a-T3b adenocarcinoma of the prostate. J. Urol., in press, 1998. 54S Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research. Loss of Cyclin-dependent Kinase Inhibitor p27Kip1 Is a Novel Prognostic Factor in Localized Human Prostate Adenocarcinoma John Tsihlias, Linda R. Kapusta, Gerrit DeBoer, et al. Cancer Res 1998;58:542-548. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/3/542 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1998 American Association for Cancer Research.