The role of p53 in the pathogenesis of, and as a predictive biomarker for, localized prostate cancer (PCa) is contested. Recent work has suggested that patterns of p53 nuclear accumulation determined by immunohistochemistry are prognostic, whereas studies using other methods question the role of p53 mutations in predicting outcome. We studied 263 men with localized PCa treated with radical prostatectomy to determine whether p53 nuclear accumulation predicts relapse and disease-specific mortality. We combined two p53 immunohistochemistry scoring systems: (a) percentage of p53-positive tumor nuclei in all major foci of cancer within the prostate; and (b) clustering, where the presence of 12 or more p53-positive cells within a ×200 power field was deemed“cluster positive.” Analysis was undertaken using χ2,Kruskal-Wallis, and Mann-Whitney tests for clinicopathological variables and the Kaplan-Meier method, log-rank test, and univariate and multivariate Cox regression modeling for evaluation of contribution to relapse and disease-specific survival. At mean follow-up of 55.1 months (range, 4.9–123.0 months), 39% (102 of 263) of patients had relapsed and 2.3% (6 of 253) had died of PCa. Pretreatment serum prostate-specific antigen concentration, pathological tumor stage,lymph node involvement, Gleason score, and p53 nuclear accumulation, as determined by either percentage score or cluster status, were independent predictors of relapse in multivariate analysis. Clustering of p53-positive cells distinguished between favorable and poor prognosis patients within the lowest p53-positive stratum (>0 to<2%) and was the most discriminatory threshold for predicting relapse in the entire cohort. p53 status predicted outcome in patients with a Gleason score of 5 and above but not those with a score of 4 and below. In patients treated with neoadjuvant hormonal therapy, p53 cluster positivity carried a 90% (19 of 21) risk of relapse by 36 months. All six patients who died from PCa in the period of the study exhibited p53 nuclear accumulation in 20% or more tumor nuclei. This study demonstrates strong relationships between p53 nuclear accumulation and relapse and disease-specific mortality in a large series of localized PCas. Furthermore, the presence of clusters of p53-positive nuclei delineates a group of patients with poor prognosis not identified by traditional scoring methods and supports the hypothesis that p53 dysfunction within PCa may exist in foci of tumor cells that are clonally expanded in metastases.

PCa4is the most common male cancer in industrialized societies and represents a serious public health problem. Identification of patients with aggressive rather than indolent PCa is a major challenge for optimal management and is only partially met by current prognostic parameters. Delineation of patterns of gene expression in early PCa that correlate with an aggressive phenotype is a priority and may allow radical treatment to be offered on a more selective basis to those patients with clinically localized yet aggressive disease.

Inactivation of the tumor suppressor gene p53 is implicated in tumorigenesis for >50% of all human cancers (1). p53 functions as a transcriptional regulator involved in G1 phase growth arrest of cells in response to DNA damage as well as having roles in the regulation of the spindle checkpoint, centrosome homeostasis, and G2-M phase transition (2). p53 also induces apoptosis by transcription-dependent and -independent mechanisms in many cell types(1, 2, 3) and regulates tumor angiogenesis and expression of the Kai1 metastasis suppressor gene (1, 4, 5, 6). Nuclear accumulation of p53 detected by IHC typically indicates the presence of p53 gene mutations (7, 8), although the correlation between nuclear accumulation of p53 and the presence of p53 gene mutations can vary (9). Nuclear accumulation of p53 is a prognostic indicator in several human cancers,including breast (4, 10, 11), lung (12), and colorectal carcinoma (13).

The value of p53 nuclear accumulation as a prognostic factor in localized PCa is controversial. A number of studies have shown that p53 nuclear accumulation detected by IHC is prognostic at a variety of dichotomizing cutoff points based on the number of p53-positive nuclei. These studies describe either a group of poor prognosis patients with 20% p53-positive nuclei (14, 15) or a group of patients with lower percentages of positive cells in a heterogeneous, focal staining pattern where either the presence of any nuclear accumulation or the presence of clusters of cells showing nuclear accumulation is adversely prognostic (16, 17, 18). However, other studies comparing p53 nuclear accumulation with assessment of p53gene mutations have failed to provide conclusive evidence for the importance of p53 in localized PCa or a strong correlation between nuclear accumulation and p53 gene mutation(19, 20, 21, 22). One study suggested that a chromosome 17p locus close to the p53 gene was an important prognostic feature when both alleles at the site were lost but found that p53 IHC is noncontributory in predicting prognosis and concluded that another gene or genes on chromosome 17p may be involved in PCa progression(19). In studying other cancers, several authors have suggested that assessment of p53 gene mutations and p53 expression in combination may more accurately define prognostically important p53 dysfunction (9, 23, 24).

Comparison of PCa metastases with primary PCas in the same patients suggest that foci with p53 mutations are clonally expanded in metastases (25, 26), perhaps explaining the high frequency of IHC positivity and the presence of gene mutations in hormone-refractory and metastatic PCa (20, 21, 22, 27, 28). Two studies have demonstrated significant heterogeneity in the distribution of p53 mutations between and within foci of carcinoma in the same prostate (29, 30). Other studies have documented heterogeneity for other genes and suggested that clones responsible for metastases do not always originate from within the dominant tumor focus (31). The possibility exists that in localized PCa, p53 overexpression and mutations as well as other genetic aberrations may be limited to subgroups of prognostically important malignant cells.

In the present study, we evaluated p53 nuclear accumulation by IHC in a series of 263 primary PCas. Nuclear accumulation was scored for overall number of positive nuclei and for the presence of clusters of 12 or more cells within a ×200 magnification field showing accumulation of p53. Assessment included sections from all identified major foci of cancer within an individual prostate. These findings were correlated with clinicopathological features, including PCa relapse and death from PCa.

Study Population.

After Research Ethics Committee approval, we studied 278 prostates from patients treated with radical retropubic prostatectomy for clinically localized adenocarcinoma of the prostate between May 1989 and December 1995 at St. Vincent’s Hospital in Sydney, Australia. Staging was undertaken by clinical examination, transrectal ultrasound, pelvic computerized tomography, and whole-body bone scanning. The 278 cases for study were selected from 409 RPs undertaken during this period based on the availability of archival tissue and were not stratified for known preoperative or pathological prognostic factors. Fifteen patients were excluded either because of a lack of a sufficient cancer to allow adequate study with contiguous sections (4 patients),inadequate follow-up information (5 patients), and recurrent technical problems related to IHC that made interpretation inconsistent, e.g. lifting of the section on antigen retrieval (6 patients). Analysis showed no statistical difference for pretreatment serum PSA, Gleason score, pathological stage, or treatment given between the 263 patients studied and the remainder of patients treated in the same time period.

Of these 263 patients, 164 were treated with RP alone with no neoadjuvant or adjuvant therapy of any kind. Thirty-nine patients received preoperative NHT (34 with goserelin and flutamide, 4 with flutamide alone, 1 with cyproterone alone) for between 1 and 8 months. Sixty-nine patients had postoperative adjuvant therapy with hormonal therapy, radiation therapy, or both (see Table 1).

Patients were followed postoperatively by their surgeons on a monthly basis until satisfactory urinary continence was obtained and then at 3-month intervals until the end of the first year, at 6-month intervals to 5 years, and yearly thereafter. Each visit included history and physical examination including DRE and measurement of serum PSA concentration using a Hybritech assay. The duration of follow-up ranged from 4.9 to 123.0 months (mean, 55.5 months), with 90% of the patients having follow-up for at least 36 months. Loss to follow-up was defined as no clinical contact and serum PSA test for >12 months at the censure date in February 1999, but no patients fulfilled this definition. Relapse was defined by the criteria: biochemical disease progression with a serum PSA concentration ≥0.4 ng/ml rising over a 3-month period; local recurrence on DRE confirmed by biopsy or by subsequent rise in PSA; or institution of long-term hormonal therapy or orchidectomy. All patients fulfilling the last criterion commenced long-term hormonal therapy or were treated with orchidectomy based on pathological features in the prostatectomy specimen.

Pathological Examination.

RP specimens were step sectioned at 2-mm intervals and completely embedded in paraffin after fixation in neutral-buffered 10% formalin. Up to three (mean, 1.74; median, 2) blocks from each case were obtained to provide representative material from the major foci of cancer within each prostate. The mean number of cancer foci sampled in this way for each case was 1.96 (median, 2; range, 1–4). The pathological staging as per the TNM classification, involvement of surgical margins, Gleason score, and WHO classification for nuclear grade and glandular differentiation was reported contemporaneously by one of three histopathologists (J. F. F., W. D., or J. J. T.) and confirmed at review with consensus where reported parameters differed (32, 33, 34).

IHC.

IHC was performed on formalin-fixed paraffin-embedded blocks sectioned at 5 μm, mounted on SuperFrost Plus slides (Menzel-Glaser,Braunschweig, Germany) and processed within 10 days of sectioning. The mouse monoclonal antibody DO-7 (DAKO Corporation,Carpinteria, CA) and avidin-biotin-peroxidase and diaminobenzidine kits (Vector Laboratories, Burlingham, CA) were used according to the manufacturers’ instructions. Briefly, sections were deparaffinized in xylene, rehydrated through graded ethanol, and then heated in a pressure cooker in 0.01 m citrate buffer (pH 6.0) for 10 min to enhance antigen retrieval. The sections were then treated with 2% H2O2 for 10 min at room temperature to inactivate endogenous peroxidase activity. After a blocking step with 10% normal horse serum, the sections were incubated with DO-7 antibody diluted to 1:200 in 2% BSA/PBS overnight at 4°C. Subsequently, sections were sequentially incubated with a biotinylated horse antimouse IgG, avidin-biotinylated complex, and diaminobenzidine. Counterstaining was undertaken with Whitlock’s hematoxylin and light green before dehydration through graded ethanol and xylene and coverslipping. A contiguous section was stained with H&E. Positive controls for p53 nuclear accumulation used with each run of staining included a paraffin-embedded pellet of the PCa cell line DU145 (35), which has a documented p53mutation; a colon cancer specimen with p53 missense mutation; and a tongue cancer specimen with p53 nuclear accumulation. Negative controls included a paraffin-embedded pellet of the PCa cell line PC3 (35), which does not express p53 protein; and the above described positive controls processed with the substitution of a non-immune mouse monoclonal antibody for the DO-7 antibody.

Scoring for p53 nuclear accumulation required assessment of all cancer in selected sections from an individual patient. Counting of a minimum of 200 cancer cells in each cancer (mean, 812; range, 210-2000 cells)was undertaken to determine the percentage of nuclei showing accumulation across all areas of cancer present. The target cell count was 500 per case where possible, but cancers with fewer cells were not excluded because of the potential for selection bias. Where the cancers had multiple foci, or were extensive or heterogeneous, more cells were counted and selected from areas of varying p53 nuclear accumulation to provide a sample representative of staining across the entirety of the cancer. Additionally, assessors scored the cancers cluster positive if 12 or more cancer cell nuclei within any ×200 power microscopic field showed p53 accumulation. In arriving at the cluster definition, we initially relied on observations of other workers using clusters of 15 cells (18, 25) and tested a variety of thresholds for the definition of a cluster within a range of 6–20 cells in the p53 score stratum with >0 to <2% p53 nuclear positivity (see below). Below 10 cells per ×200 field, there was no correlation with prognosis, and positive cells tended to be dispersed throughout the cancer rather than clustered within the area of a single field, whereas there were few cases with clusters of >15 cells within the >0 to <2% p53 score stratum. This definition differentiated for outcome when between 10 and 15 cells (log-rank, P = 0.04 and P = 0.05, respectively, within the >0 to<2% stratum, whereas use of 9 cells produced P = 0.12, and 16 cells produced P = 0.24)were included as thresholds; therefore 12 cells (log-rank, P < 0.001) within the field was selected as a midpoint in this range. Sections were scored independently for p53 by two assessors (D. I. Q. and S. M. H.) and one pathologist (J. F. F., W. D., or J. J. T.),all of whom were blinded to patient outcome. The interobserver Spearman rank coefficients for p53 score were between 0.92 and 0.96, signifying close agreement between scorers. The χ2 test for the p53 cluster status initially assigned by different scorers produced P values of 0.87 and 0.76. Specifically, the two assessors identified 12 p53 cluster-positive cases not identified on initial assessment by the pathologist, whereas the pathologist identified 2 cases not initially identified by the assessors. All of these cases were deemed cluster positive at consensus review.

Statistical Analysis.

Data were evaluated for relapse and disease-specific mortality prediction using the Kaplan-Meier product limit method and log-rank test, and by univariate and multivariate analysis in a Cox proportional hazards model for p53 cluster status and score, and other recognized clinical and pathological predictors of outcome (36, 37). To produce multivariate models relevant to clinical practice, variables including factors previously described as predictive of outcome and found to be statistically significant on univariate analysis, but excluding p53 status, were modeled as dichotomized or continuous variables to determine their independent prognostic value. Further modeling with independent variables and p53 score was then undertaken. All statistical analyses were performed using StatView 4.5 software(Abacus Systems, Berkeley, CA). Statistical significance in this study was set at P < 0.05. All reported P values are two-sided.

Clinical and Pathological Characteristics.

The mean age of the patients at surgery was 63.2 years (range,43.8–76.7 years). The mean preoperative serum PSA concentration was 19.7 ng/ml (n = 260; range, 1.0–280 ng/ml;median, 11.7 ng/ml). The original diagnostic indications for prostatectomy were as follows: unknown, n = 2(0.8%); restaging of PCa diagnosed on transurethral resection, n = 18 (6.8%); abnormal DRE alone, n = 27 (10.3%); PSA elevation alone, n = 93 (35.5%); and abnormal DRE with PSA elevation, n = 122 (46.6%). The distribution of PCa clinical stage according to the TNM classification is presented in Table 1.

Of 263 patients, 115 had organ-confined disease; 139 had extraprostatic extension, 36 had SVI; 142 had no surgical margin involvement, whereas 62 had a single margin positive and 58 had multiple positive margins; and 5 had pelvic lymph node metastases. The pathological stage and correlation with a series of pathological variables are shown in Table 2. Forty-five (17%) patients had well-differentiated, 151 (57%) had moderately differentiated, and 67 (26%) had poorly differentiated tumors according to the WHO classification (34).

Clinical, treatment, and outcome data for the cohort are summarized in Table 1. One hundred two (38.8%) patients relapsed during the period of follow-up (mean, 55.1 months; range, 4.9–123.0 months). Mean time to relapse was 19.6 months (range, 0.1–59.5 months). Ninety-two patients who relapsed developed elevated serum PSA measurements in the period of study either as the first sign of relapse or subsequent to evidence of local recurrence. Twelve patients developed evidence of local recurrence either with or without PSA elevation as the first sign of recurrence, but all developed a simultaneous or subsequent rise in PSA. Ten patients were deemed to have relapsed on the basis of commencement of continuous postoperative hormonal therapy or orchidectomy based on adverse features within the histopathology report. Omission of these 10 patients from the analyses described below yielded results essentially the same as for the overall cohort in each instance. Fourteen patients in the cohort died during the study period,6 of them from PCa.

Significant predictors of relapse on univariate analysis in this series of clinically localized PCas treated with RP were preoperative serum PSA, clinical stage, Gleason score, worst single Gleason grade,surgical margin involvement, overall pathological stage, pathological tumor stage, extraprostatic extension, SVI, and lymph node involvement(see Tables 3 and 4). Patients undergoing NHT or any form of adjuvant therapy had a significantly worse prognosis compared with patients treated with RP alone (see Tables 3 and 4).

Multivariate analysis was initially undertaken using representative prognostic variables of univariate significance and excluding those patients who commenced long-term hormonal therapy or who had orchidectomy performed based on the results of histopathology because these patients were deemed to have relapsed at institution of such therapy. This analysis showed that NHT (P = 0.33) and adjuvant therapy (P = 0.1) become nonsignificant when other predictive variables were introduced, as did surgical margin involvement (P = 0.46) and SVI (P = 0.09; Table 4).

p53 Nuclear Accumulation and Relapse.

The pattern of p53 nuclear accumulation seen was consistent with that described in previous studies where there was considerable heterogeneity between different areas of cancer within the same prostate and within single foci of carcinoma. Homogeneous p53 nuclear accumulation was unusual, with heterogeneous staining containing foci of varying size encompassing a variable number of p53-positive cells being the commonest pattern (Fig. 1). The p53 percentage score and p53 cluster status were significantly correlated with Gleason score, pathological stage, and pretreatment serum PSA concentration (Table 2). Univariate analysis demonstrated that p53 score as a continuous variable was prognostic (Table 3).

Kaplan-Meier product limit analysis for relapse demonstrated differences in relapse for strata based on the percentage of cells showing p53 nuclear accumulation at the levels of 0%, >0–2%,>2–5%, >5–20%, and >20% positive cells, respectively (Fig. 2 A; overall log-rank, P < 0.0001). Differences between individual strata were significant by log-rank calculation between 0 and >0 to <2% (P = 0.01), >0 to <2% and ≥2 to <5% (P = 0.0002), and ≥2–5% and ≥20% (P = 0.009).

p53 Clustering and Relapse.

The use of p53 cluster status produced the most discriminatory demarcation point for relapse in the cohort when analyzed by Kaplan-Meier and univariate analysis (see Fig. 3, A and B, and Tables 3 and 4).

p53 cluster status was discriminatory for cases with a Gleason score≥5 (log-rank, P < 0.0001) but not for those with a score ≤4 (log-rank, P = 0.61), and for those cases designated moderately (log-rank, P < 0.0001) and poorly differentiated (log-rank, P < 0.0001) on WHO criteria but not for those designated well differentiated (log-rank, P = 0.28). p53 cluster status was statistically discriminatory for relapse in all pathological staging groups including cases where the cancer was organ confined(pT2N0; log-rank, P < 0.0001).

In comparing the separate scoring systems, all cancers with p53 nuclear accumulation >2% were also p53 cluster positive. Those cancers with<2% p53 nuclear accumulation exhibited the presence of clusters in a proportion of cases with no strict relationship to the overall percentage of tumor nuclei positive, i.e. a single cluster could be present as the only p53-positivity in the entire cancer. Given this observation, the stratum for a p53 nuclear accumulation score between 0 and 2% was further subdivided according to the presence or absence of clusters. Subsequently, Kaplan-Meier analysis demonstrated that cluster-positive cases within this stratum relapsed at a similar rate to those with ≥2 to <5% p53 nuclear accumulation (Fig. 1,B; log-rank, P = 0.73; and Fig. 1,C; log rank, P = 0.46), and this was confirmed by univariate analysis (Tables 3 and 4) and maintained in multivariate models (Table 5). There was no significant difference in outcome between those cases deemed cluster negative within the >0 to <2% stratum and those in the 0 stratum (log-rank, P = 0.19), although those patients within the 0 stratum had significantly better relapse-free survival compared with all other patients, i.e. those with any p53 nuclear accumulation (log-rank, P < 0.0001).

When representative independently prognostic variables were then modeled with p53 cluster status, p53 score as a continuous variable,and p53 strata, p53 nuclear accumulation proved to be an independent prognostic indicator of relapse (Table 5). When stepwise multivariate analyses were constructed with backward elimination in the Cox proportional hazard regression model, the factors most predictive of relapse were, in descending order: lymph node involvement, Gleason score, p53 status measured by either score or cluster status,pretreatment PSA less than or greater than 10 ng/ml, and pathological stage pT3 or greater against pT2.

p53 Nuclear Accumulation and Relapse in Patients Treated with NHT.

Thirty-nine patients received NHT prior to RP, 23 of whom relapsed, and 1 of whom died of PCa during the study period. p53 score (Mann-Whitney, P = 0.0008), p53 strata(χ2, P = 0.005), and p53 cluster status (Fisher’s exact test, P = 0.0002) predicted relapse in the NHT group. Ninety percent (19 of 21)of NHT patients who were p53 cluster positive relapsed within 36 months of prostatectomy (Fig. 4; log-rank, P < 0.0001) compared with 22% (4 of 18) of those who were p53 cluster negative. Univariate analysis in this group revealed that p53 cluster status and pretreatment PSA concentration were significant predictors of outcome (Table 6). Multivariate analysis extended to include Gleason score stratified at the 4–7 and 8–10 levels and pathological T stage stratified between pT2 and pT3 or greater showed that all were significant predictors of outcome in the model(Table 6). However, with stepwise regression analysis, the factors most predictive of relapse were, in descending order: p53 cluster positivity, pretreatment PSA level, Gleason score of 8–10, and pathological T stage. In bivariate analysis using each of these variables with p53 cluster status, pretreatment PSA level maintained statistical significance, whereas Gleason score and pathological T stage did not. Hence, p53 cluster status was the strongest predictor of outcome in patients treated with NHT prior to RP.

p53 Nuclear Accumulation and Survival.

Six patients died from PCa at a mean of 52.3 months (range, 13.5–104.4 months) following surgery. These patients had a mean time to PSA relapse of 10.9 months (range, 0.5–37.0 months) compared with 20.2 months (range, 0.1–59.5 months) in the group of patients that had relapsed but not died of PCa (P = 0.14). The mean time from surgery to development of clinical or bone scan-detected metastatic disease in the six patients to die from PCa was 28.7 months(range, 8–88 months). All patients that died from PCa had a p53 score of ≥20%. Kaplan-Meier analysis of the ≥20% stratum against the<20% strata was highly significant in predicting death from PCa (log-rank, P < 0.0001; Fig. 4). Evaluation of p53 status in predicting overall survival failed to reach significance (data not shown).

Our study illustrates a strong relationship between p53 nuclear accumulation detected by IHC in clinically localized PCa and relapse and PCa-related death in a large well-characterized cohort. Univariate analysis for clinicopathological factors described as predicting outcome following RP demonstrates that the cohort has characteristics consistent with previous large studies of outcome (Refs.38, 39, 40, 41, and Table 3). The novel findings of this study are as follows: that increasing p53 score carries with it an increased risk of relapse and death; that the presence of a cluster of p53-positive tumor cells within a ×200 magnification field provides a discriminatory point predictive of relapse; and that p53 nuclear accumulation is the most powerful predictor of outcome in patients given NHT prior to RP. These data support the findings of previous studies that demonstrated relationships between outcome and p53 IHC at thresholds of no nuclear accumulation versus any accumulation (16, 17), the presence of clusters of p53-positive cells versus their absence (18, 42), and between ≥20% p53-positive nuclei and <20%(14, 15). In predicting early death from PCa following RP,p53 positivity in ≥20% of nuclei defines a group of patients with highly aggressive disease that progresses much more rapidly than is the case for most patients experiencing PSA relapse (41). Our data also support previously described relationships between p53 nuclear accumulation and pathological tumor stage and Gleason score,and define a positive correlation between p53 score and pretreatment serum PSA concentration. These data run counter to a number of studies that found p53 nuclear accumulation to be uncommon and/or nonprognostic in localized PCa (19, 20, 21, 22).

p53 status was not predictive of outcome in patients with better differentiated tumors, i.e. those that were well differentiated or with a Gleason score ≤4. The outcome following surgery for patients with better differentiated tumors was good, and there was a low event rate in these groups. It may be that an effect of p53 clustering has not been apparent within our follow-up period. Alternatively, it may be that these cancers are intrinsically indolent and that p53 status is not of importance within this subset because of a lack of other genetic and epigenetic factors contributing to cancer progression. Conversely, p53 status predicted outcome in all subgroups based on pathological stage and pretreatment serum PSA strata. One potential limitation of our study is that although biochemical relapse correlates with subsequent development of clinical metastases, the rate of development of these metastases varies greatly; further evaluation of this and other cohorts is needed to determine whether p53 status predicts clinical relapse.

p53 cluster status was the most predictive factor in determining outcome in patients who received NHT prior to RP (Table 6). Grignon et al.(15), in reporting on 129 patients with clinically localized PCa treated with radiation therapy in the RTOG 8610 trial, found that p53 nuclear accumulation in pretreatment diagnostic material predicted reduced time to distant metastases in those patient given NHT but not in those treated with radiation therapy alone. The predictive value of p53 IHC in the NHT group raises interesting questions about the effect of hormonal therapy on PCa cells with and without p53 nuclear accumulation. Following androgen ablation,a small proportion of PCa cells undergo apoptosis, but the majority of cells responding undergo cell cycle arrest mediated, at least in part,by p53, entering G0 phase and losing cell volume(43, 44). Cells with dysfunctional p53 may be resistant to hormonal therapy and fail to undergo cell cycle arrest or apoptosis(15, 45), conferring a relative growth advantage and greater prominence when evaluated for p53 nuclear accumulation. Such a postulated mechanism may explain the observation that p53 dysfunction is associated with hormonal resistance in some studies of breast cancer and PCa (15, 45).

Our study reports a higher rate of p53 positivity than some previous studies, with 79% of cases showing at least occasional p53 positivity and 52% being p53 cluster positive (Table 2). The most obvious explanation for this relates to a modified definition of p53 positivity that integrates p53 cluster status into the p53 accumulation score. Another potential reason is the inclusion of a greater proportion of patients with adverse features not as prevalent in other series, such as pathological stage (57%pT3N0 or greater), tumor differentiation (16% with Gleason score ≥8 and 26% poorly differentiated by WHO criteria), and pretreatment serum PSA concentration (25% greater than 20 ng/ml), that contribute to a relapse rate of 38.8% at the mean follow-up of 55.1 months. In a study of 175 RP cases evaluated for p53 nuclear accumulation, 65% of cases were reported to demonstrate occasional or greater (i.e. >0) p53 positivity (17) with a corresponding relapse of 37.7% at a mean follow-up of 55.2 months. Within that series(17), 57.1% of cases were pathological stage pT3N0 or greater, 10.8%had a Gleason score ≥8, and 5.7% were poorly differentiated;pretreatment serum PSA concentrations were not reported. This suggests that although the stage of tumors studied was similar in both studies,the tumors in our study had a higher Gleason score, were more poorly differentiated on WHO criteria, and had a slightly higher chance of relapse. The inclusion of patients receiving neoadjuvant and adjuvant therapy is likely to have resulted in a cohort with relatively more aggressive or advanced cancer as indicated by pathological stage,Gleason score, and preoperative serum PSA concentrations. Further comparison between our cohort and a recently reported large single-surgeon case series, where 54.4% of cases were pathological stage pT3N0 or greater,7.8% had a Gleason score ≥8, and 5.5% of patients had a pretreatment serum PSA concentration ≥20 ng/ml, also suggests that our group contains patients with more adverse features than others(41). Given these differences and the trend toward screening-detected cancers treated with RP being of lower stage,further evaluation of p53 nuclear accumulation in cohorts with less aggressive features is desirable. Finally, it may be that the antigen retrieval and IHC techniques used in our laboratory are more sensitive than those reported in some other studies.

Methodological factors may also have contributed to our higher p53-positivity rate because of the assessment of all major foci within each individual prostate and the use of a p53 antibody directed at the DO7 epitope, which has a high correlation with p53 gene mutation compared with other antibodies (46). Other researchers have found that a cocktail of DO1 and DO7 epitope-directed antibodies is slightly more sensitive than DO7 alone (47). We achieved equivalent results with DO7 antibody, which detects both wild-type and mutant p53 protein, having noted increased background staining in initial experiments with the described cocktail (data not shown). In determining p53 status, it appears that IHC is sensitive but may suffer from lack of specificity for detecting mutation, whereas methods of direct sequencing are highly specific in detecting p53 mutation but may lack sensitivity when such mutations are present in only a small percentage of cells within a population containing wild-type p53. This may in part explain differences in p53 status based on the technique used.

Accumulation of p53 protein may occur in response to a number of stimuli independent of p53 gene mutation, including DNA damage, hypoxia, and redox stress (48). Elledge et al.(49, 50) have suggested that even low levels of p53 protein accumulation are prognostic in breast cancer specimens regardless of whether p53 gene mutation could be detected concurrently, whereas Silvestrini et al.(51)have shown that an increasing percentage of p53-positive nuclei between 0 and 12% has a corresponding adverse “dose” effect on prognosis in a large series of breast cancer patients. Our study describes a similar relationship between percentage of p53-positive nuclei and relapse in PCa, but it also illustrates that clusters of p53-positive cells are prognostic when <2% of all tumor cells are positive. The prognostic significance of p53 clusters suggests that cells in such clusters have important characteristics such as p53 mutation or are affected by other factors capable of producing local p53 nuclear accumulation and are associated with a poorer outcome. In this regard,p53 clusters may represent foci of cells with p53 gene mutations that expand within the prostate to increase the p53 score with time so that more advanced and/or higher Gleason grade cancer is associated with a higher p53 score. It also is possible that micrometastases occur at a similar time in PCa progression to the development of clusters with a critical number of cells showing p53 nuclear accumulation. This hypothesis suggests a focal dose-response threshold for p53 dysfunction and genesis of metastases (see below).

PCa is usually multifocal. The clone responsible for human PCa metastases may reside in smaller tumor foci within the prostate gland rather than the largest (31, 52, 53). Heterogeneity between and within different foci of PCa within the same gland is inferred by studies of DNA flow cytometry (54) and allelic loss (52, 55). Heterogeneity of p53 mutations and p53 protein nuclear accumulation within and between individual foci of early-stage localized cancer in the same prostate has been demonstrated previously (29, 30). p53 is crucial in determining metastatic potential in models of skin carcinogenesis where p53 protein dose does not contribute to initiation and/or promotion but is associated with metastasis (56). Similarly, one study found that in ras+myc-initiated prostate carcinoma metastasis is concurrent with loss of expression of the wild-type p53 allele (57). In that study,comparative DNA analysis between primary tumors and metastases demonstrated that metastases did not necessarily derive from the most abundant clone but seeded from small subpopulations from within the primary tumor (57). The clonal expansion of p53 dysfunctional cells (57) has been confirmed in small series of primary human PCas and metastases within the same patients(25, 26, 58). These studies add to others that demonstrate increased p53 nuclear accumulation in metastatic, recurrent, and/or androgen-insensitive PCa compared with clinically localized disease(20, 21, 22, 27, 28). In a study of 50 metastatic PCa foci in 19 men with lethal PCa, p53 gene mutations demonstrated homogeneity for mutation in virtually all metastases assessed from individual patients, in contrast to PTEN/MMAC1 mutations,which demonstrated intermetastasis heterogeneity (59). Taken together, these studies suggest that prostate tumor cells harboring p53 mutations and perhaps other genetic aberrations are clonally expanded in metastases. Our study extends this concept by suggesting that a “p53 dose” effect across the entire cancer and at a given focal threshold within clusters is important in the metastatic process.

The work presented in this study and that of others (16, 17, 18, 60) suggests that p53 has an important prognostic role in approximately half of all patients with clinical localized PCa. Given the finding that p53 is the most important predictor of outcome in patients given NHT before RP and similar observations in patients given NHT prior to radiation therapy (15), p53 status may have significant implications for patients treated with hormonal therapy as part of these regimens. The genetic and epigenetic basis for clustered p53 nuclear accumulation and its effect on clinical as well as biochemical relapse require further investigation.

Fig. 1.

Representative photomicrographs demonstrating p53 nuclear accumulation in PCa. Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody with hematoxylin and light green counterstain (see “Materials and Methods”). A,cluster-positive case with Gleason 4 PCa acinar formation with >12 cells demonstrating p53 nuclear accumulation (brown;×400 magnification). B, PCa Gleason single grade 4 with no nuclei demonstrating p53 nuclear accumulation (×200 magnification). C, PCa Gleason single grade 4 with >70% of nuclei demonstrating p53 nuclear accumulation (×200 magnification).

Fig. 1.

Representative photomicrographs demonstrating p53 nuclear accumulation in PCa. Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody with hematoxylin and light green counterstain (see “Materials and Methods”). A,cluster-positive case with Gleason 4 PCa acinar formation with >12 cells demonstrating p53 nuclear accumulation (brown;×400 magnification). B, PCa Gleason single grade 4 with no nuclei demonstrating p53 nuclear accumulation (×200 magnification). C, PCa Gleason single grade 4 with >70% of nuclei demonstrating p53 nuclear accumulation (×200 magnification).

Close modal
Fig. 2.

A, relapse-free survival for patients with clinical localized PCa treated with RP with or without neoadjuvant or adjuvant therapy categorized by p53 IHC score into strata: 0 (×);>0 to <2% (♦); ≥2 to <5% (▿); ≥5 to <20% (⋄); ≥20%(▴) of immunoreactive nuclei. Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method (n = 263; overall log-rank, P < 0.0001;intergroup log-rank: 0 versus >0 to <2%, P = 0.01; >0 to <2%versus ≥2 to <5%, P = 0.002; ≥2 to <5% versus ≥5% to <20%, P = 0.24; and ≥5% to <20%versus >20%, P = 0.19). B, relapse-free survival for patients with clinically localized PCa treated with RP with or without neoadjuvant or adjuvant therapy categorized by p53 IHC strata incorporating p53 cluster status:0 (×); >0 to <2% and cluster negative (▪); >0 to <2% and cluster positive (▾); ≥2 to <5% (▿), ≥5 to <20% (⋄); ≥20%() of immunoreactive nuclei (n = 263, overall log-rank, P < 0.0001; intergroup log-rank: 0 versus >0 to <2%,cluster negative, P = 0.19; >0 to <2%,cluster negative versus >0 to <2%, cluster positive, P = 0.0004; >0 to <2%, cluster positive versus ≥2 to <5%, P = 0.73; ≥2 to <5% versus ≥5% to <20%, P = 0.24; ≥5% to <20%versus ≥20%, P = 0.19;and ≥2 to <5% versus ≥20%, P = 0.009). C,relapse-free survival for patients with localized PCa treated with RP alone by p53 IHC strata as for B, demonstrating a similar relationship between p53 score and relapse in this subset of patients (n = 164; overall log-rank, P < 0.0001; all intergroup log-rank P values were not significant except that for >0 to<2%, cluster negative versus >0 to <2%, cluster positive: P = 0.048).

Fig. 2.

A, relapse-free survival for patients with clinical localized PCa treated with RP with or without neoadjuvant or adjuvant therapy categorized by p53 IHC score into strata: 0 (×);>0 to <2% (♦); ≥2 to <5% (▿); ≥5 to <20% (⋄); ≥20%(▴) of immunoreactive nuclei. Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method (n = 263; overall log-rank, P < 0.0001;intergroup log-rank: 0 versus >0 to <2%, P = 0.01; >0 to <2%versus ≥2 to <5%, P = 0.002; ≥2 to <5% versus ≥5% to <20%, P = 0.24; and ≥5% to <20%versus >20%, P = 0.19). B, relapse-free survival for patients with clinically localized PCa treated with RP with or without neoadjuvant or adjuvant therapy categorized by p53 IHC strata incorporating p53 cluster status:0 (×); >0 to <2% and cluster negative (▪); >0 to <2% and cluster positive (▾); ≥2 to <5% (▿), ≥5 to <20% (⋄); ≥20%() of immunoreactive nuclei (n = 263, overall log-rank, P < 0.0001; intergroup log-rank: 0 versus >0 to <2%,cluster negative, P = 0.19; >0 to <2%,cluster negative versus >0 to <2%, cluster positive, P = 0.0004; >0 to <2%, cluster positive versus ≥2 to <5%, P = 0.73; ≥2 to <5% versus ≥5% to <20%, P = 0.24; ≥5% to <20%versus ≥20%, P = 0.19;and ≥2 to <5% versus ≥20%, P = 0.009). C,relapse-free survival for patients with localized PCa treated with RP alone by p53 IHC strata as for B, demonstrating a similar relationship between p53 score and relapse in this subset of patients (n = 164; overall log-rank, P < 0.0001; all intergroup log-rank P values were not significant except that for >0 to<2%, cluster negative versus >0 to <2%, cluster positive: P = 0.048).

Close modal
Fig. 3.

Relapse-free survival for patients with clinically localized PCa treated with RP categorized by p53 IHC cluster status:“positive” refers to those cancers with a cluster of immunoreactive nuclei within the strata >0 to <2%, and all patients with >2% of nuclei scored positive; “negative” refers to patients in the strata>0 to <2% without clusters as well as those patients with 0 nuclei showing immunoreactivity. ▪, negative; ▾, positive. Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method. A,all patients treated with RP with or without other therapy.(n = 263; P < 0.0001). B, patients treated with RP alone(n = 164; P = 0.0001). C, patients treated with NHT followed by RP (n = 39; P = 0.0001).

Fig. 3.

Relapse-free survival for patients with clinically localized PCa treated with RP categorized by p53 IHC cluster status:“positive” refers to those cancers with a cluster of immunoreactive nuclei within the strata >0 to <2%, and all patients with >2% of nuclei scored positive; “negative” refers to patients in the strata>0 to <2% without clusters as well as those patients with 0 nuclei showing immunoreactivity. ▪, negative; ▾, positive. Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method. A,all patients treated with RP with or without other therapy.(n = 263; P < 0.0001). B, patients treated with RP alone(n = 164; P = 0.0001). C, patients treated with NHT followed by RP (n = 39; P = 0.0001).

Close modal
Fig. 4.

Disease-specific survival for patients with clinically localized PCa treated with RP categorized by p53 IHC score <20% (+)or ≥20% (×) of tumor nuclei showing p53 accumulation: Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method(n = 263; P < 0.0001).

Fig. 4.

Disease-specific survival for patients with clinically localized PCa treated with RP categorized by p53 IHC score <20% (+)or ≥20% (×) of tumor nuclei showing p53 accumulation: Survival curves were generated according to the Kaplan-Meier method, and statistical comparisons were made by use of the log-rank method(n = 263; P < 0.0001).

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This research was supported by grants from the National Health and Medical Research Council of Australia, RT Hall Trust, Laurence Freedman Trust, New South Wales Cancer Council, Leo and Jenny Leukemia, and Cancer Foundation of Australia, St. Vincent’s Clinic Foundation, and Merck Sharp and Dohme Research Foundation. D. Q. is a National Health and Medical Research Council of Australia Medical Postgraduate Research Scholar and recipient of the Vincent Fairfax Family Foundation Fellowship from the Royal Australasian College of Physicians.

4

The abbreviations used are: PCa, prostate cancer; IHC, immunohistochemistry; PSA, prostate-specific antigen; RP,radical prostatectomy; NHT, neoadjuvant hormonal therapy; DRE, digital rectal examination; TNM, Tumor-Node-Metastasis; SVI, seminal vesicle involvement.

Table 1

Clinical, treatment, and outcome data for 263 patients with clinically localized PCa treated with RP

CharacteristicNumberPercentage
Age (n = 263)   
>65 years 119 43.5 
Clinical stagea (n = 263)   
T1A 2.7 
T1B 14 5.3 
T1C 74 28.1 
T2A 73 27.8 
T2B 59 22.4 
T2C 23 8.7 
T3 14 5.3 
Treatment (n = 263)b   
RP alone 164 62.4 
NHT 39 14.8 
Total adjuvant therapy 69 26.2 
Adjuvant radiation therapy 15 5.7 
Adjuvant hormonal therapy 46 17.5 
Adjuvant radiation therapy and hormonal therapy 2.7 
Continuous postoperative hormonal therapyc 10 3.8 
Relapsed 102 38.8 
Death from PCa 2.3 
Death from any cause 14 5.3 
CharacteristicNumberPercentage
Age (n = 263)   
>65 years 119 43.5 
Clinical stagea (n = 263)   
T1A 2.7 
T1B 14 5.3 
T1C 74 28.1 
T2A 73 27.8 
T2B 59 22.4 
T2C 23 8.7 
T3 14 5.3 
Treatment (n = 263)b   
RP alone 164 62.4 
NHT 39 14.8 
Total adjuvant therapy 69 26.2 
Adjuvant radiation therapy 15 5.7 
Adjuvant hormonal therapy 46 17.5 
Adjuvant radiation therapy and hormonal therapy 2.7 
Continuous postoperative hormonal therapyc 10 3.8 
Relapsed 102 38.8 
Death from PCa 2.3 
Death from any cause 14 5.3 
a

Tumor stage according to TNM (32).

b

Nine NHT patients received adjuvant therapy (4 radiation therapy, 5 hormonal therapy ≤3 months), and therefore, total adds to more than 100%.

c

Orchidectomy or continuous postoperative hormonal therapy(defined as therapy intended to be long-term) and undertaken or commenced on the basis of pathologic characteristics of the PCa at RP.

d

Relapse defined as serum PSA concentration at or above 0.4 ng/ml, rising over a 3-month period; local recurrence on DRE confirmed by biopsy or by subsequent rise in PSA; or institution of long-term hormonal therapy or orchidectomy.

Table 2

Association of p53 immunoreactivity with tumor Gleason score, stage,and pretreatment PSA levels

No. of subjectsp53 nuclear accumulation strataa (% of patients within each group)p53 clusteraMean p53 scoreb
0>0 to <2%, cluster negative>0 to <2%, cluster positive≥2 to <5%≥5 to <20%≥20%P% positivePP
All  263 20.5 27.8 12.2 13.3 11.0 12.9  50.2  7.4  
Gleason scorec (n = 258) 2–4  23 39.1 26.1 4.3 13.0 13.0 4.3  34.8  2.8  
 5–7  194 17.0 29.4 16.5 11.9 10.8 11.3 0.02 47.9 0.009 6.7 0.0007 
 8–10 41 4.9 24.4 9.8 22.0 12.2 26.8  70.7  13.8  
Pathological staged (n = 263) PT2N0 114 28.1 29.8 12.2 11.4 8.8 9.7  43.0  4.7  
 PT3N0 108 19.4 29.6 15.7 14.8 13.9 6.5  50.0  4.9  
 PT3CN0 29 13.8 24.1 3.4 27.6 6.9 24.1 <0.0001 62.1 0.01 13.7 <0.0001 
 PT4N0 14.3 85.7  85.7  43.6  
 PTXN+ 40 60  100.0  36.0  
Pretreatment PSA concentration (n = 260)e <4 14 35.7 21.4 0.0 21.4 0.0 21.4  42.8  12.9  
 4–10 96 28.1 39.6 8.3 6.3 10.4 7.3  32.3  4.2  
 10.1–20 86 17.4 24.4 17.4 12.8 14.0 14.0 0.003 58.1 <0.0001 8.0 0.007 
 >20 64 15.6 17.2 14.1 25.0 10.9 17.2  67.2  9.9  
No. of subjectsp53 nuclear accumulation strataa (% of patients within each group)p53 clusteraMean p53 scoreb
0>0 to <2%, cluster negative>0 to <2%, cluster positive≥2 to <5%≥5 to <20%≥20%P% positivePP
All  263 20.5 27.8 12.2 13.3 11.0 12.9  50.2  7.4  
Gleason scorec (n = 258) 2–4  23 39.1 26.1 4.3 13.0 13.0 4.3  34.8  2.8  
 5–7  194 17.0 29.4 16.5 11.9 10.8 11.3 0.02 47.9 0.009 6.7 0.0007 
 8–10 41 4.9 24.4 9.8 22.0 12.2 26.8  70.7  13.8  
Pathological staged (n = 263) PT2N0 114 28.1 29.8 12.2 11.4 8.8 9.7  43.0  4.7  
 PT3N0 108 19.4 29.6 15.7 14.8 13.9 6.5  50.0  4.9  
 PT3CN0 29 13.8 24.1 3.4 27.6 6.9 24.1 <0.0001 62.1 0.01 13.7 <0.0001 
 PT4N0 14.3 85.7  85.7  43.6  
 PTXN+ 40 60  100.0  36.0  
Pretreatment PSA concentration (n = 260)e <4 14 35.7 21.4 0.0 21.4 0.0 21.4  42.8  12.9  
 4–10 96 28.1 39.6 8.3 6.3 10.4 7.3  32.3  4.2  
 10.1–20 86 17.4 24.4 17.4 12.8 14.0 14.0 0.003 58.1 <0.0001 8.0 0.007 
 >20 64 15.6 17.2 14.1 25.0 10.9 17.2  67.2  9.9  
a

Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody (see “Materials and Methods”). Stratification of tumors was based on percentage of tumor cell nuclei demonstrating immunoreactivity. Cases where tumor cells demonstrated>0 but <2% nuclei immunoreactivity were further subdivided for the presence or absence of clusters of 12 or more immunoreactive nuclei with a ×200 magnification field (cluster negative or cluster positive): 0%; >0 to <2% and cluster negative; >0 to <2% and cluster positive, ≥2 to <5%; ≥5 to <20%; and ≥20% of immunoreactive nuclei.

b

Mean p53 score as a percentage of tumor cells with p53 nuclear accumulation. For the purpose of calculation of a mean p53 score for each group, p53 scores <1% were assumed to be 0.5%.

c

Tumor grade according to Gleason criteria (33).

d

Tumor stage according to TNM (32).

e

PSA concentrations as described in “Materials and Methods.”

Table 3

Univariate analysis for clinicopathologic and biological variables with disease-free survival following RP in 263 patients with clinically localized PCa

VariableHazards ratio95% confidence intervalP
Age >65 vs. <65 years 0.79 0.53–1.18 0.25 
NHT plus RP vs. RP alone 2.07 1.19–3.60 0.01 
Adjuvant therapy vs. no adjuvant therapy (n = 253)a 1.63 1.05–2.53 0.03 
Clinical stageb    
T1 1.00   
T2 1.51 0.96–2.35 0.71 
T3 4.25 2.06–8.76 0.003 
Pretreatment PSAc,d (n = 260) 1.012 1.008–1.017 <0.0001 
Pretreatment PSA >10 vs. <10 ng/mlc 3.05 1.93–4.82 <0.0001 
pT stageb    
pT2 1.00   
pT3 2.08 1.29–3.34 0.003 
pT3C 3.92 2.19–6.99 0.02 
pT4 11.16 5.20–23.91 0.03 
SVI 3.08 1.97–4.81 <0.0001 
Lymph node involvement 58.84 19.97–173.41 <0.0001 
Margins    
Nil 1.00   
Single margin 1.64 1.01–2.66 0.04 
Multiple margins 2.65 1.67–4.20 0.09 
Pathological stagec    
pT2N0 1.00   
pT3N0 2.02 1.26–3.23 0.004 
pT3CN0 3.50 1.93–6.33 0.47 
pT4N0 7.61 3.13–18.49 0.17 
pTXN+ 108.76 35.13–336.69 0.023 
Gleason gradede (n = 258) 1.59 1.38–1.83 <0.0001 
p53 scored,f 1.033 1.024–1.041 <0.0001 
p53 score strataf    
1.00   
>0 to <2%, N 1.99 0.71–5.57 0.19 
>0 to <2%, P 7.08 2.59–19.34 0.004 
≥2 to <5% 7.88 2.97–20.92 0.9 
≥5 to <20% 11.03 4.13–29.44 0.17 
≥20% 16.63 6.39–43.28 0.2 
p53 cluster status negative vs. positivef 6.42 3.85–10.70 <0.0001 
VariableHazards ratio95% confidence intervalP
Age >65 vs. <65 years 0.79 0.53–1.18 0.25 
NHT plus RP vs. RP alone 2.07 1.19–3.60 0.01 
Adjuvant therapy vs. no adjuvant therapy (n = 253)a 1.63 1.05–2.53 0.03 
Clinical stageb    
T1 1.00   
T2 1.51 0.96–2.35 0.71 
T3 4.25 2.06–8.76 0.003 
Pretreatment PSAc,d (n = 260) 1.012 1.008–1.017 <0.0001 
Pretreatment PSA >10 vs. <10 ng/mlc 3.05 1.93–4.82 <0.0001 
pT stageb    
pT2 1.00   
pT3 2.08 1.29–3.34 0.003 
pT3C 3.92 2.19–6.99 0.02 
pT4 11.16 5.20–23.91 0.03 
SVI 3.08 1.97–4.81 <0.0001 
Lymph node involvement 58.84 19.97–173.41 <0.0001 
Margins    
Nil 1.00   
Single margin 1.64 1.01–2.66 0.04 
Multiple margins 2.65 1.67–4.20 0.09 
Pathological stagec    
pT2N0 1.00   
pT3N0 2.02 1.26–3.23 0.004 
pT3CN0 3.50 1.93–6.33 0.47 
pT4N0 7.61 3.13–18.49 0.17 
pTXN+ 108.76 35.13–336.69 0.023 
Gleason gradede (n = 258) 1.59 1.38–1.83 <0.0001 
p53 scored,f 1.033 1.024–1.041 <0.0001 
p53 score strataf    
1.00   
>0 to <2%, N 1.99 0.71–5.57 0.19 
>0 to <2%, P 7.08 2.59–19.34 0.004 
≥2 to <5% 7.88 2.97–20.92 0.9 
≥5 to <20% 11.03 4.13–29.44 0.17 
≥20% 16.63 6.39–43.28 0.2 
p53 cluster status negative vs. positivef 6.42 3.85–10.70 <0.0001 
a

Excludes 10 patients treated with orchidectomy or continuous postoperative hormonal therapy defined as therapy intended to be long-term and undertaken or commenced based on pathologic characteristics of the PCa at RP.

b

Tumor stage according to TNM (32).

c

PSA concentrations determined as described in “Materials and Methods.”

d

Calculations use continuous variables so that for each unit(ng/ml) increase in pretreatment PSA, there is a 1.2% increase in risk of relapse; for each unit increase in Gleason score, there is a 59%increase in risk of relapse; and for each percentage point increase in p53 score, there is 3.3% increase in risk of relapse. For the purpose of calculation of a mean p53 score for each group, p53 scores below 1%were assumed to be 0.5%. Note that P values presented in more than two tiered variables compare the category on the same line with the one above so that, for example, when T3N0tumors are compared to pT2N0 tumors, P = 0.004, when pT3CN0 tumors are compared with pT3N0 tumors, P = 0.47 and so on.

e

Tumor grade according to Gleason criteria (33).

f

Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody (see “Materials and Methods”). p53 score refers to the percentage of tumor cells demonstrating p53 nuclear accumulation. Stratification of tumors was based on percentage of tumor cell nuclei demonstrating immunoreactivity and, in cases where tumor cells demonstrated >0 but <2% nuclei immunoreactivity, were further subdivided for the presence or absence of clusters of 12 or more immunoreactive nuclei within a ×200 magnification field [cluster negative (N) or cluster positive (P)]: 0; >0 to <2% and cluster negative; >0 to <2% and cluster positive; ≥2 to <5%, ≥5 to<20%; and ≥20% of immunoreactive nuclei.

Table 4

Univariate and multivariate analysis of described prognostic factors and treatment as dichotomized or continuous variables in determining outcome (n = 245)a

VariableUnivariate analysisMultivariate analysis
Hazards ratio (95% confidence interval)PHazards ratio (95% confidence interval)P
NHT plus RP vs. RP alone 2.07 (1.19–3.60) 0.01 1.36 (0.73–2.52) 0.33 
Adjuvant therapy vs. no adjuvant therapy 1.63 (1.05–2.53) 0.03 0.63 (0.36–1.09) 0.10 
Clinical stageb T3vs. T1 and T2 3.25 (1.70–6.26) 0.0004 1.22 (0.51–2.91) 0.66 
Pretreatment PSAc >10 vs. <10 ng/ml 3.05 (1.93–4.82) <0.0001 2.52 (1.55–4.09) 0.0002 
Tumor stageb pT2 vs. pT3 or greater 2.67 (1.72–4.14) <0.0001 1.75 (1.02–3.01) 0.04 
SVI positive vs. negative 3.08 (1.97–4.81) <0.0001 1.62 (0.92–2.84) 0.09 
Lymph node positive vs. negative 54 (19.97–173) <0.0001 10.88 (1.26–93.94) 0.03 
Margins multiple vs. nil & single 1.87 (1.18–2.96) 0.008 1.22 (0.71–2.08) 0.46 
Gleason scored 1.59 (1.38–1.83) <0.0001 1.40 (1.17–1.68) 0.0003 
VariableUnivariate analysisMultivariate analysis
Hazards ratio (95% confidence interval)PHazards ratio (95% confidence interval)P
NHT plus RP vs. RP alone 2.07 (1.19–3.60) 0.01 1.36 (0.73–2.52) 0.33 
Adjuvant therapy vs. no adjuvant therapy 1.63 (1.05–2.53) 0.03 0.63 (0.36–1.09) 0.10 
Clinical stageb T3vs. T1 and T2 3.25 (1.70–6.26) 0.0004 1.22 (0.51–2.91) 0.66 
Pretreatment PSAc >10 vs. <10 ng/ml 3.05 (1.93–4.82) <0.0001 2.52 (1.55–4.09) 0.0002 
Tumor stageb pT2 vs. pT3 or greater 2.67 (1.72–4.14) <0.0001 1.75 (1.02–3.01) 0.04 
SVI positive vs. negative 3.08 (1.97–4.81) <0.0001 1.62 (0.92–2.84) 0.09 
Lymph node positive vs. negative 54 (19.97–173) <0.0001 10.88 (1.26–93.94) 0.03 
Margins multiple vs. nil & single 1.87 (1.18–2.96) 0.008 1.22 (0.71–2.08) 0.46 
Gleason scored 1.59 (1.38–1.83) <0.0001 1.40 (1.17–1.68) 0.0003 
a

Excludes patients treated with continuous postoperative hormonal therapy or adjuvant orchidectomy.

b

Tumor stage according to TNM (32).

c

PSA concentrations determined as described in “Materials and Methods.”

d

Tumor grade according to Gleason criteria (33). Calculations use Gleason score as a continuous variable so that for each unit increase in Gleason score there is a 40% increase in risk of relapse.

Table 5

Multivariate analyses incorporating prognostic variables of significance within the model tested in Table 4 modeled with p53 score,p53 strata, and p53 cluster status (n = 255)

VariableHazards ratio (95% confidence interval)
p53 score modelp53 strata modelp53 cluster model
Pretreatment PSAa <10, >10 ng/ml 2.54 (1.60–4.05) 2.11 (1.32–3.38) 2.01 (1.25–3.21) 
Tumor stageb pT2vs. pT3 or greater 1.76 (1.10–2.81) 1.78 (1.10–2.87) 1.67 (1.03–2.69) 
Lymph node negative vs. positive 13.51 (3.81–47.86) 15.88 (5.12–49.21) 19.97 (6.66–59.88) 
Gleason scorec 1.35 (1.16–1.57) 1.33 (1.13–1.55) 1.38 (1.18–1.62) 
p53 scored 1.022 (1.012–1.032)   
p53 stratad    
0%  1.00  
>0 to <2%; cluster negative  1.67 (0.60–4.70)  
>0 to <2%; cluster positive  5.12 (1.87–14.06)  
≥2 to <5%  4.94 (1.84–13.28)  
≥5 to <20%  7.55 (2.78–20.41)  
≥20%  10.44 (3.93–27.74)  
p53 clusterd positive vs. negative   4.67 (2.76–7.88) 
VariableHazards ratio (95% confidence interval)
p53 score modelp53 strata modelp53 cluster model
Pretreatment PSAa <10, >10 ng/ml 2.54 (1.60–4.05) 2.11 (1.32–3.38) 2.01 (1.25–3.21) 
Tumor stageb pT2vs. pT3 or greater 1.76 (1.10–2.81) 1.78 (1.10–2.87) 1.67 (1.03–2.69) 
Lymph node negative vs. positive 13.51 (3.81–47.86) 15.88 (5.12–49.21) 19.97 (6.66–59.88) 
Gleason scorec 1.35 (1.16–1.57) 1.33 (1.13–1.55) 1.38 (1.18–1.62) 
p53 scored 1.022 (1.012–1.032)   
p53 stratad    
0%  1.00  
>0 to <2%; cluster negative  1.67 (0.60–4.70)  
>0 to <2%; cluster positive  5.12 (1.87–14.06)  
≥2 to <5%  4.94 (1.84–13.28)  
≥5 to <20%  7.55 (2.78–20.41)  
≥20%  10.44 (3.93–27.74)  
p53 clusterd positive vs. negative   4.67 (2.76–7.88) 
a

PSA concentrations determined as described in“Materials and Methods.”

b

Tumor stage according to TNM (32).

c

Tumor grade according to Gleason criteria (33). Calculations use Gleason score as a continuous variable so that for each unit increase in Gleason score, there is a 33–38% increase in risk of relapse.

d

Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody (see “Materials and Methods”). For details of p53 score, p53 score strata, and p53 cluster status, see Table 3 and“Materials and Methods.” Calculations use p53 score as a continuous variable so that for each percentage point increase in p53 staining,the risk of relapse increases by 2.2%.

Table 6

Univariate and multivariate analysis of prognostic factors in determining relapse-free survival of 39 patients treated with NHT prior to RP

ParameterUnivariate analysisMultivariate analysis
Hazards ratio95% confidence intervalPHazards ratio95% confidence intervalP
pT stagea       
pT2 1.00   1.00   
pT3 or greater 2.57 0.87–7.57 0.087 4.17 0.95–18.35 0.06 
Gleason scoreb       
4–7 1.00   1.00   
8–10 3.43 0.76–15.56 0.11 22.06 2.70–180.13 0.007 
Pretreatment PSAc 1.025 1.005–1.014 0.0012 1.020 1.004–1.036 0.016 
p53 clusterd       
Negative 1.00   1.00   
Positive 6.98 2.30–21.17 0.0006 6.35 2.05–19.69 0.0014 
ParameterUnivariate analysisMultivariate analysis
Hazards ratio95% confidence intervalPHazards ratio95% confidence intervalP
pT stagea       
pT2 1.00   1.00   
pT3 or greater 2.57 0.87–7.57 0.087 4.17 0.95–18.35 0.06 
Gleason scoreb       
4–7 1.00   1.00   
8–10 3.43 0.76–15.56 0.11 22.06 2.70–180.13 0.007 
Pretreatment PSAc 1.025 1.005–1.014 0.0012 1.020 1.004–1.036 0.016 
p53 clusterd       
Negative 1.00   1.00   
Positive 6.98 2.30–21.17 0.0006 6.35 2.05–19.69 0.0014 
a

Tumor stage according to TNM (32).

b

Tumor grade according to Gleason criteria (33).

c

PSA concentrations determined as described in “Materials and Methods.”

d

Immunohistochemical detection of p53 nuclear accumulation was performed using 5-μm thick sections of tumor tissue and DO-7 monoclonal p53 antibody and assessed for the presence of p53 clusters(see “Materials and Methods”).

1
Kirsch D. G., Kastan M. B. Tumor-suppressor p53: implications for tumor development and prognosis.
J. Clin. Oncol.
,
16
:
3158
-3168,  
1998
.
2
Agarwal M. L., Taylor W. R., Chernov M. V., Chernova O. B., Stark G. R. The p53 network.
J. Biol. Chem.
,
273
:
1
-4,  
1998
.
3
Liebermann D. A., Hoffman B., Steinman R. A. Molecular controls of growth arrest and apoptosis: p53-dependent and independent markers.
Oncogene
,
11
:
199
-210,  
1995
.
4
Gasparini G., Weidner N., Bevilacqua P., Maluta S., Dalla Palma P., Caffo O., Barbareschi M., Boracchi P., Marubini E., Pozza F. Tumor microvessel density, p53 expression, tumor size, and peritumoral lymphatic vessel invasion are relevant prognostic markers in node-negative breast carcinoma.
J. Clin. Oncol.
,
12
:
454
-466,  
1994
.
5
Yu E. D., Yu E., Meyer G. E., Brawer M. K. The relation of p53 protein nuclear accumulation and angiogenesis in human prostate cancer.
Prostate Cancer Prostate Dis.
,
1
:
39
-44,  
1997
.
6
Mashimo T., Watabe M., Hirota S., Hosobe S., Miura K., Tegtmeyer P. J., Rinker-Shaeffer C. W., Watabe K. The expression of the KAI1 gene, a tumor metastasis suppressor, is directly activated by p53.
Proc. Natl. Acad. Sci. USA
,
95
:
11307
-11311,  
1998
.
7
Finlay C. A., Hinds P. W., Tan T. H., Eliyahn D., Oren M., Levine A. J. Activating mutations for transformation by p53 produce a gene product that forms an hsc70–p53 complex with an altered half-life.
Mol. Cell. Biol.
,
8
:
531
-539,  
1988
.
8
Bartek J., Iggo R., Gannon J., Lane D. P. Genetic and immunochemical analysis of mutant p53 in human breast cancer cell lines.
Oncogene
,
5
:
893
-899,  
1990
.
9
Casey G., Lopez M. E., Ramos J. C., Plummer S. J., Arboleda M. J., Shaughnessy M., Karlan B., Slamon D. J. DNA sequence analysis of exons 2 through 11 and immunohistochemical staining are required to detect all known p53 alterations in human malignancies.
Oncogene
,
13
:
1971
-1981,  
1996
.
10
Thor A. D., Moore D. H., Edgerton S. M., Kawasaki E. S., Reihaus E., Lynch H. T., Marcus J. N., Schwartz L., Chen L-C., Mayall B. H., Smith H. S. Accumulation of p53 tumor suppressor gene protein: an independent marker of prognosis in breast cancers.
J. Natl. Cancer Inst.
,
84
:
845
-855,  
1992
.
11
Elledge R. M., Clark G. M., Fuqua S. A., Yu Y. Y., Allred D. C. p53 protein accumulation detected by five different antibodies: relationship to prognosis and heat shock protein 70 in breast cancer.
Cancer Res.
,
54
:
3752
-3757,  
1994
.
12
Xu H-J., Cagle P. T., Hu S-X., Li J., Benedict W. F. Altered retinoblastoma and p53 protein status in non-small cell carcinoma of the lung: potential synergistic effects on prognosis.
Clin. Cancer Res.
,
2
:
1169
-1176,  
1996
.
13
Goh H-S., Yao J., Smith D. R. p53 point mutation and survival in colorectal cancer patients.
Cancer Res.
,
55
:
5217
-5221,  
1995
.
14
Visakorpi T., Kallioniemi O. P., Heikkinen A., Koivula T., Isola J. Small subgroup of aggressive, highly proliferative prostatic carcinomas defined by p53 accumulation.
J. Natl. Cancer Inst.
,
84
:
883
-887,  
1992
.
15
Grignon D. J., Caplan R., Sarkar F. H., Lawton C. A., Hammond E. H., Pilepich M. V., Forman J. D., Mesic J., Fu K. K., Abrams R. A., Pajak T. F., Shipley W. U., Cox J. D. p53 status and prognosis of locally advanced prostatic adenocarcinoma: a study based on RTOG 8610.
J. Natl. Cancer Inst.
,
89
:
158
-165,  
1997
.
16
Bauer J. J., Sesterhenn I. A., Mostofi K. F., McLeod D. G., Srivastava S., Moul J. W. p53 nuclear protein expression is an independent prognostic marker in clinically localized prostate cancer patients undergoing radical prostatectomy.
Clin. Cancer Res.
,
1
:
1295
-1300,  
1995
.
17
Bauer J. J., Sesterhenn I. A., Mostofi F. K., McLeod D. G., Srivastava S., Moul J. W. Elevated levels of apoptosis regulator proteins p53 and bcl-2 are independent prognostic biomarkers in surgically treated clinically localized prostate cancer.
J. Urol.
,
156
:
1511
-1516,  
1996
.
18
Yang G., Stapleton A. M. F., Wheeler T. M., Truong L. D., Timme T. L., Scardino P. T., Thompson T. C. Clustered p53 immunostaining: a novel pattern associated with prostate cancer progression.
Clin. Cancer Res.
,
2
:
399
-401,  
1996
.
19
Brooks J. D., Bova G. S., Ewing C. M., Piantadosi S., Carter B. S., Robinson J. C., Epstein J. I., Isaacs W. B. An uncertain role for p53 gene alterations in human prostate cancers.
Cancer Res.
,
56
:
3814
-3822,  
1996
.
20
Navone N. M., Troncoso P., Pisters L. L., Goodrow T. L., Palmer J. L., Nichols W. W., von Eschenbach A. C., Conti C. J. p53 protein accumulation and gene mutation in the progression of human prostate carcinoma.
J. Natl. Cancer Inst.
,
85
:
1657
-1669,  
1993
.
21
Bookstein R., MacGrogan D., Hilsenbeck S. G., Sharkey F., Allred D. C. p53 is mutated in a subset of advanced-stage prostate cancers.
Cancer Res.
,
53
:
3369
-3373,  
1993
.
22
Dinjens W. N., van der Weiden M. M., Schroeder F. H., Bosman F. T., Trapman J. Frequency and characterization of p53 mutations in primary and metastatic human prostate cancer.
Int. J. Cancer
,
56
:
630
-633,  
1994
.
23
Chi S. G., de Vere White R. W., Meyers F. J., Siders D. B., Lee F., Gumerlock P. H. p53 in prostate cancer: frequent expressed transition mutations.
J. Natl. Cancer Inst.
,
86
:
926
-933,  
1994
.
24
Gumerlock P. H., Chi S. G., Shi X. B., Voeller H. J., Jacobson J. W., Gelmann E. P., de Vere White R. W. p53 abnormalities in primary prostate cancer: single-strand conformation polymorphism analysis of complementary DNA in comparison with genomic DNA. The Cooperative Prostate Network.
J. Natl. Cancer Inst.
,
89
:
66
-71,  
1997
.
25
Stapleton A. M. F., Timme T. L., Gousse A. E., Li Q-F., Tobon A. A., Kattan M. W., Slawin K. M., Wheeler T. M., Scardino P. T., Thompson T. C. Primary human prostate cancer cells harboring p53 mutations are clonally expanded in metastases.
Clin. Cancer Res.
,
3
:
1389
-1397,  
1997
.
26
Meyers F. J., Gumerlock P. H., Chi S. G., Borchers H., Deitch A. D., de Vere White R. W. Very frequent p53 mutations in metastatic prostate carcinoma and in matched primary tumors.
Cancer (Phila.)
,
83
:
2534
-2539,  
1998
.
27
Prendergast N. J., Atkins M. R., Schatte E. C., Paulson D. F., Walther P. J. p53 immunohistochemical and genetic alterations are associated at high incidence with post-irradiated locally persistent prostate carcinoma.
J. Urol.
,
155
:
1685
-1692,  
1996
.
28
Grossfeld G. D., Olumi A. F., Connolly J. A., Chew K., Gibney J., Bhargava V., Waldman F. M., Carroll P. R. Locally recurrent prostate tumors following either radiation therapy or radical prostatectomy have changes in Ki-67 labeling index, p53 and bcl-2 immunoreactivity.
J. Urol.
,
159
:
1437
-1443,  
1998
.
29
Konishi N., Hiasa Y., Matsuda H., Tao M., Tsuzuki T., Hayashi I., Kitahori Y., Shiraishi T., Yatani R., Shimazaki J. Intratumor cellular heterogeneity and alterations in ras oncogene and p53 tumor suppressor gene in human prostate carcinoma.
Am. J. Pathol.
,
147
:
1112
-1122,  
1995
.
30
Mirchandani D., Zheng J., Miller G. J., Ghosh A. K., Shibata D. K., Cote R. J., Roy-Burman P. Heterogeneity in intratumor distribution of p53 mutations in human prostate cancer.
Am. J. Pathol.
,
147
:
92
-101,  
1995
.
31
Jenkins R. B., Qian J., Lieber M. M., Bostwick D. G. Detection of c-myc oncogene amplification and chromosomal anomalies in metastatic prostatic carcinoma by fluorescence in situ hybridization.
Cancer Res.
,
57
:
524
-531,  
1997
.
32
Schroder F. H., Hermanek P., Denis L., Fair W. R., Gospodarowicz M. K., Pavone-Macaluso M. The TNM classification of prostate cancer.
Prostate
,
4
:
129
-138,  
1992
.
33
Gleason D. F. Histologic grading and clinical staging of prostatic carcinoma Tannebaum M. eds. .
Urologic Pathology: The Prostate
,
:
171
-197, Lea and Febiger Philadelphia  
1977
.
34
Mostofi, F. K., Sesterhenn, I. A., and Sobin, L. H. Histological typing of prostate tumours. Geneva: WHO, 1980.
35
Carroll A. G., Voeller H. J., Sugars L., Gelmann E. P. p53 oncogene mutations in three human prostate cancer cell lines.
Prostate
,
23
:
123
-134,  
1993
.
36
Kaplan E. L., Meier P. Nonparametric estimation from incomplete observations.
J. Am. Stat. Assoc.
,
53
:
457
-481,  
1958
.
37
Cox D. R. Regression models and life tables (life tables).
J. R. Stat. Soc.
,
34
:
187
-189,  
1972
.
38
Ohori M., Wheeler T. M., Kattan M. W., Goto Y., Scardino P. T. Prognostic significance of positive surgical margins in radical prostatectomy specimens.
J. Urol.
,
154
:
1818
-1824,  
1995
.
39
Zincke H., Bergstralh E. J., Blute M. L., Myers R. P., Barrett D. M., Lieber M. M., Martin S. K., Oesterling J. E. Radical prostatectomy for clinically localized prostate cancer: long-term results of 1143 patients from a single institution.
J. Clin. Oncol.
,
12
:
2254
-2263,  
1994
.
40
Cher M. L., Shinohara K., Breslin S., Vapnek J., Carroll P. R. High failure rate associated with long-term follow-up of neoadjuvant androgen deprivation followed by radical prostatectomy for stage C prostatic cancer.
Br. J. Urol.
,
75
:
771
-777,  
1995
.
41
Pound C. R., Partin A. W., Eisenberger M. A., Chan D. W., Walsh P. C. Natural history of progression after PSA elevation following radical prostatectomy.
JAMA
,
281
:
1591
-1597,  
1999
.
42
Stapleton A. M., Zbell P., Kattan M. W., Yang G., Wheeler T. M., Scardino P. T., Thompson T. C. Assessment of the biologic markers p53, Ki-67, and apoptotic index as predictive indicators of prostate carcinoma recurrence after surgery.
Cancer (Phila.)
,
82
:
168
-175,  
1998
.
43
Westin P., Stattin P., Damber J. E., Bergh A. Castration therapy rapidly induces apoptosis in a minority and decreases cell proliferation in a majority of human prostatic tumors.
Am. J. Pathol.
,
146
:
1368
-1375,  
1995
.
44
Agus D. B., Cordon-Cardo C., Fox W., Drobnjak M., Koff A., Golde D. W., Scher H. I. Prostate cancer cell cycle regulators: response to androgen withdrawal and development of androgen independence.
J. Natl. Cancer Inst.
,
91
:
1869
-1876,  
1999
.
45
Bergh J., Norberg T., Sjogren S., Lindgren A., Holmberg L. Complete sequencing of the p53 gene provides prognostic information in breast cancer patients, particularly in relation to adjuvant systemic therapy and radiotherapy.
Nat. Med.
,
1
:
1029
-1034,  
1995
.
46
Thomas M. D., McIntosh G. G., Anderson J. J., McKenna D. M., Parr A. H., Johnstone R., Lennard T. W., Horne C. H., Angus B. A novel quantitative immunoassay system for p53 using antibodies selected for optimum designation of p53 status.
J. Clin. Pathol.
,
50
:
143
-147,  
1997
.
47
Wertz I. E., Deitch A. D., Gumerlock P. H., Gandour-Edwards R., Chi S. G., de Vere White R. W. Correlation of genetic and immunodetection of TP53 mutations in malignant and benign prostate tissues.
Hum. Pathol.
,
27
:
573
-580,  
1996
.
48
Giaccia A. J., Kastan M. B. The complexity of p53 modulation: emerging patterns from divergent signals.
Genes Dev.
,
12
:
2973
-2983,  
1998
.
49
Elledge R. M., Gray R., Mansour E., Yu Y., Clark G. M., Ravdin P., Osborne C. K., Gilchrist K., Davidson N. E., Robert N. Accumulation of p53 protein as a possible predictor of response to adjuvant combination chemotherapy with cyclophosphamide, methotrexate, fluorouracil, and prednisone for breast cancer.
J. Natl. Cancer Inst.
,
87
:
1254
-1256,  
1995
.
50
Elledge R. M. Assessing p53 status in breast cancer prognosis: where should you put the thermometer if you think your p53 is sick?.
J. Natl. Cancer Inst.
,
88
:
141
-143,  
1996
.
51
Silvestrini R., Daidone M. G., Benini E., Faranda A., Tomasic G., Boracchi P., Salvadori B., Veronesi U. Validation of p53 accumulation as a predictor of distant metastasis at 10 years of follow up in 1400 node-negative breast cancers.
Clin. Cancer Res.
,
2
:
2007
-2013,  
1996
.
52
Sakr W. A., Macoska J. A., Benson P., Grignon D. J., Wolman S. R., Pontes J. E., Crissman J. D. Allelic loss in locally metastatic, multisampled prostate cancer.
Cancer Res.
,
54
:
3273
-3277,  
1994
.
53
Geburek B. M., Kollmorgen T. A., Qian J., D’Souza-Gburek S. M., Lieber M. M., Jenkins R. B. Chromosomal anomalies in stage D1 prostate adenocarcinoma primary tumors and lymph node metastases detected by fluorescence in situ hybridization.
J. Urol.
,
157
:
223
-227,  
1997
.
54
O’Malley F. P., Grignon D. J., Keeney M., Kerkvliet N., McLean C. DNA heterogeneity in prostatic adenocarcinoma. A DNA flow cytometric mapping study with whole organ sections of prostate.
Cancer (Phila.)
,
71
:
2797
-2802,  
1993
.
55
Cheng L., Song S. Y., Pretlow T. G., Abdul-Karim F. W., Kung H. J., Dawson D. V., Park W. S., Moon Y. W., Tsai M. L., Linehan W. M., Emmert-Buck M. R., Liotta L. A., Zhuang Z. Evidence of independent origin of multiple tumors from patients with prostate cancer.
J. Natl. Cancer Inst.
,
90
:
233
-237,  
1998
.
56
Kemp C. J., Donehower L. A., Bradley A., Balmain A. Reduction of p53 gene dosage does not increase initiation or promotion but enhances malignant progression of chemically induced skin tumors.
Cell
,
74
:
813
-822,  
1993
.
57
Thompson T. C., Park S. H., Timme T. L., Ren C., Eastham J. A., Donehower L. A., Bradley A., Kadmon D., Yang G. Loss of p53 function leads to metastasis in ras+myc-initiated mouse prostate cancer.
Oncogene
,
10
:
869
-879,  
1995
.
58
Navone N. M., Labate M. E., Troncoso P., Pisters L. L., Conti C. J., von Eschenbach A. C., Logothetis C. J. p53 mutations in prostate cancer bone metastases suggest that selected p53 mutants in the primary site define foci with metastatic potential.
J. Urol.
,
161
:
304
-308,  
1999
.
59
Suzuki H., Freije D., Nusskern D. R., Okami K., Cairns P., Sidransky D., Isaacs W. B., Bova G. S. Interfocal heterogeneity of PTEN/MMAC1 gene alterations in multiple metastatic prostate cancer tissues.
Cancer Res.
,
58
:
204
-209,  
1998
.
60
Brewster S. F., Oxley J. D., Trivella M., Abbott C. D., Gillatt D. A. Preoperative p53, bcl-2, CD44, and E-cadherin immunohistochemistry as predictors of biochemical relapse after radical prostatectomy.
J. Urol.
,
161
:
1238
-1243,  
1999
.