Elevated osteonectin/SPARC expression in primary prostate cancer predicts metastatic progression.

BACKGROUND: The majority of prostate cancers (CaP) are detected in early stages with uncertain prognosis. Therefore, an intensive effort is underway to define early predictive markers of CaP with aggressive progression characteristics. METHODS: In order to define such prognostic markers, we performed comparative analyses of transcriptomes of well- and poorly differentiated (PD) tumor cells from primary tumors of patients (N=40) with 78 months of mean follow-up after radical prostatectomy. Validation experiments were carried out at transcript level by quantitative real-time reverse transcriptase-PCR (RT-PCR) (N=110) and at protein level by immunohistochemistry (N=53) in primary tumors from an independent patient cohort. RESULTS: Association of a biochemical network of 12 genes with SPARC gene as a central node was highlighted with PD phenotype. Of note, there was remarkable enrichment of NKXH_NKXH_HOX composite regulatory elements in the promoter of the genes in this network suggesting a biological significance of this gene-expression regulatory mechanism in CaP progression. Further, quantitative expression analyses of SPARC mRNA in primary prostate tumor cells of 110 patients validated the association of SPARC expression with poor differentiation and higher Gleason score. Most significantly, higher SPARC protein expression at the time of prostatectomy was associated with the subsequent development of metastasis (P=0.0006, AUC=0.803). CONCLUSIONS: In summary, we propose that evaluation of SPARC in primary CaP has potential as a prognostic marker of metastatic progression.

Overexpression of C-MYC oncogene in prostate cancer predicts biochemical recurrence.

Alterations of chromosome 8, including amplification at 8q24 harboring the C-MYC oncogene, have been noted as one of the most common chromosomal abnormalities in prostate cancer (CaP) progression. However, the frequency of C-MYC alterations in CaP has remained uncertain. A recent study, using a new anti-MYC antibody, described prevalent upregulation of nuclear C-MYC protein expression as an early oncogenic alteration in CaP. Further, we have recently reported regulation of C-MYC expression by ERG and a significant correlation between C-MYC overexpression and TMPRSS2-ERG fusion in early stage CaP. These emerging data suggest that increased C-MYC expression may be a critical and early oncogenic event driving CaP progression. In this study, we assessed whether C-MYC mRNA overexpression in primary prostate tumors was predictive of more aggressive tumor or disease progression. Our approach was to quantitatively determine C-MYC mRNA expression levels in laser capture micro-dissected tumor cells and matched benign epithelial cells in a radical prostatectomy cohort with long follow-up data available. On the basis of our results, we conclude that elevated C-MYC expression in primary prostate tumor is biologically relevant and may be a predictor of future biochemical recurrence.

ERG oncoprotein expression in prostate cancer: clonal progression of ERG-positive tumor cells and potential for ERG-based stratification.

Gene fusions prevalent in prostate cancer (CaP) lead to the elevated expression of the ERG proto-oncogene. ERG activation present in 50-70% of prostate tumors underscores one of the most common oncogenic alterations in CaP. Despite numerous reports of gene fusions and mRNA expression, ERG oncoprotein status in CaP still remains to be defined. Furthermore, development of ERG protein-based assays may provide a new dimension to evaluation of gene fusions involving diverse androgen-regulated promoters and the ERG protein-coding sequence. Through exhaustive evaluations of 132 whole-mount prostates (261 tumor foci and over 200 000 benign glands) for the ERG oncoprotein nuclear expression, we demonstrated 99.9% specificity for detecting prostate tumor cells using a highly specific anti-ERG monoclonal antibody. The ERG oncoprotein expression correlated well with fusion transcript or gene fusion in randomly selected specimens. Strong concordance of ERG-positive foci of prostatic intraepithelial neoplasia (PIN) with ERG-positive carcinoma (82 out of 85 sections with PIN, 96.5%) affirms the biological role of ERG in clonal selection of prostate tumors in 65% (86 out of 132) of patients. Conversely, ERG negative PINs were associated with ERG-negative carcinoma. Taken together, the homogeneous and strong ERG expression detected in individual tumors establishes the potential for ERG oncoprotein-based stratification of CaP.

Quantitative expression of TMPRSS2 transcript in prostate tumor cells reflects TMPRSS2-ERG fusion status.

TMPRSS2-ERG fusion is the most common oncogenic rearrangement in prostate cancer (CaP). Owing to this chromosomal rearrangement one TMPRSS2 allele loses its promoter, and one of the ETS-related gene (ERG) alleles gains that promoter leading to its overexpression in these tumor cells. Some studies suggest that TMPRSS2, an androgen-regulated type II transmembrane serine protease, may have an effect on CaP progression. We hypothesized that a difference in TMPRSS2 expression may be present in vivo between CaP cells with and without TMPRSS2-ERG fusion, or a compensatory mechanism for the allelic loss of TMPRSS2 may balance that expression difference. Therefore, TMPRSS2 mRNA expression was evaluated in microdissected CaP cells with and without TMPRSS2-ERG fusion in 132 CaP patients and analyzed for its correlation with other androgen receptor (AR)-regulated genes and clinicopathological features. In vivo TMPRSS2 expression correlated with that of other AR-regulated genes, including PSA/KLK3 and PMEPA1, offering potential as AR surrogates. A significantly reduced expression of TMPRSS2 was evident in malignant cells harboring TMPRSS2-ERG fusion, but not in CaP cells without TMPRSS2-ERG fusion, further defining these two genetically distinct types of CaP.

TMPRSS2-ERG fusion, a common genomic alteration in prostate cancer activates C-MYC and abrogates prostate epithelial differentiation.

The high prevalence of TMPRSS2-ERG rearrangements ( approximately 60%) in prostate cancer (CaP) leads to androgenic induction of the ETS-related gene (ERG) expression. However, the biological functions of ERG overexpression in CaP remain to be understood. ERG knockdown in TMPRSS2-ERG expressing CaP cells induced striking morphological changes and inhibited cell growth both in cell culture and SCID mice. Evaluation of the transcriptome and specific gene promoters in ERG siRNA-treated cells and investigation of gene expression signatures of human prostate tumors revealed ERG-mediated activation of C-MYC oncogene and the repression of prostate epithelial differentiation genes (PSA and SLC45A3/Prostein). Taken together, these data combining cell culture and animal models and human prostate tumors reveal that ERG overexpression in prostate tumor cells may contribute to the neoplastic process by activating C-MYC and by abrogating prostate epithelial differentiation as indicated by prostate epithelial specific markers.

Do patients with low volume prostate cancer have prostate specific antigen recurrence following radical prostatectomy?

AIMS: The objective of this study was to determine the incidence of prostate specific antigen (PSA) relapse in patients with low volume prostate cancer following radical prostatectomy. METHODS: Between 1993 and 2001, 50 of 717 patients had total tumour volumes of less than 0.5 cm3 following radical prostatectomy. Biochemical recurrence was defined as two consecutive values of serum PSA levels of 0.2 ng/ml or greater. RESULTS: Median follow-up of the 50 patients was 58 months. In five of the 50 patients (10%), PSA recurrence was observed. All of these five cases had Gleason score of 3+3 (well and/or moderately differentiated), organ confined and surgical margin negative tumours. In three of the five cases, capsular incision resulted in benign glands extending into the surgical margin. CONCLUSIONS: Five of 50 cases had PSA failure. In three of the five patients, benign glands located in the margin could explain the "PSA recurrence". However, in the other two patients, none of the pathological parameters correlated with measurable PSA levels. The explanation for their PSA failure is unclear.

Transcriptome analyses of benign and malignant prostate epithelial cells in formalin-fixed paraffin-embedded whole-mounted radical prostatectomy specimens.

Formalin-fixed paraffin-embedded (FFPE) prostate specimens are rich sources of molecular pathological information. However, FFPE-based microarray analysis of tissue samples may be hampered by the degradation and chemical alteration of RNA molecules due to the preservation procedure. In this report, we employed a probe analyses of Affymetrix oligonucleotide arrays at individual probe level to compensate for the potential loss of gene identifications associated with compromised mRNA quality in FFPE preparations. Furthermore, to increase the sample quality, we utilized laser capture microdissection of prostate tumor and benign epithelial cells. Remarkably, combination of these approaches recapitulated the common prostate cancer-associated gene expression alteration. Identification of prostate cancer associated-gene expression alterations such as AMACR, Kallikrein gene family and genes associated with androgen signaling such as PDEF and STEAP were consistent with previous findings reported in prostate cancer. These data suggest that combination of laser capture dissection with computational enhancement of microarray data may be useful for the assessment of gene expression changes in FFPE prostate cancer specimens.

Androgen receptor mutation (T877A) promotes prostate cancer cell growth and cell survival.

Alteration of the AR functions due to amplification, overexpression and somatic mutation of the AR itself or altered interaction of AR with other cell growth regulatory proteins, may contribute to a significant subset of advanced prostate cancer (CaP). Very little is known about the pathways impacted by AR dysfunctions, although CaP associated AR alterations suggest the biological role of the AR dysfunction in disease progression. Comparative evaluations of wild type (wt) AR and mutant (mt) ARs in appropriate experimental models should provide a better understanding of the functional impact of AR alterations in CaP. Here, we provide direct evidence showing cell growth/cell survival promoting effects of the widely studied CaP associated AR mutation (T877A). In contrast to Ad-wtAR or Ad-control infected LNCaP or LAPC4 cells, Ad-mtAR (T877A) infected LNCaP or LAPC4 cells continued to grow in the androgen-deprived medium and exhibited an androgen independent AR-transcription factor activity. Further, Ad-mtAR (T877A) infected LNCaP or LAPC4 cells exhibited enhanced cell growth in the presence of lower concentrations of the synthetic androgen, R1881. Of note, Ad-mtAR (T877A) infected LNCaP cells showed striking resistance to cell growth inhibition/apoptosis mediated by the wt p53. Taken together, these findings provide novel insights into the AR dysfunctions resulting from the T877A mutation and functionally similar AR alterations may provide selective cell growth/survival advantage for CaP progression. These observations have important implications for developing biology-based prognostic biomarkers and therapeutic strategies for CaP showing such AR dysfunctions.

Quantitative expression profile of PSGR in prostate cancer.

PSGR is a novel member of the G-protein-coupled olfactory receptor family. Our initial report showed predominant expression of the PSGR in human prostate gland and significant alterations of PSGR expression in primary prostate cancer (CaP) specimens. The aim of this study was to provide in-depth evaluations of the expression profile of PSGR in prostatic epithelial cells of CaP patients and to evaluate the association of PSGR expression characteristics with clinico-pathologic features. In total, 220 RNA specimens, from laser capture microdissected paired benign and malignant prostatic epithelial cells of 110 CaP patients, were analyzed for PSGR expression by quantitative real-time PCR. The differential expression of PSGR between the prostatic epithelial cells of malignant and benign glands was statistically significant (P<0.0001). Comparison of PSGR expression between paired benign and tumor cells revealed prostate tumor cell-specific overexpression in 67.2% of tumor specimens (74 of 110), decreased expression in 20.9% of tumor specimens (23 of 110) and no difference of PSGR expression between tumor and normal cells in 11.8% of specimens (13 of 110). In representative cases, PSGR expression patterns were independently confirmed by in situ RNA hybridization. The PSGR overexpression associated with higher percentage of pathologic stage, pT3, and a higher level of preoperative serum PSA. CaP cells of African-American CaP patients exhibited about two-fold increase of PSGR expression in comparison to the Caucasian American CaP patients. Strikingly high-percentage CaP cells overexpress PSGR warrants further studies of PSGR expression alterations to define subsets of CaPs.

Phenology and field biology of black cutworm (Lepidoptera: Noctuidae) in Ontario no-till corn.

Black cutworm, Agrotis ipsilon (Hufnagel), is an occasional corn, Zea mays L., pest that is attracted to no-till fields. Understanding the phenology of black cutworm in Ontario no-till corn, particularly the time of arrival of adults in relation to the onset of crop damage and the stages of larvae that coincide with vulnerable corn seedling leaf stages, is important for their effective control. Pheromone and blacklight trap captures of moths first occurred in early April, whereas significant influxes did not occur until mid- to late April. Males and females were often captured simultaneously, in contrast to findings in Iowa and Illinois where males were captured in pheromone traps on average 3 wk ahead of females or males in blacklight traps. This may be a reflection of a more mature source population for the influxes into Ontario because first captures also were later than in the United States. Females arrived mated and corn seedling cutting occurred within 137 degree-days (DD) (base 10.4 degrees C) of first capture in Ontario corn. Cutworms were present in cornfields before planting, and the mean age of larvae increased along with corn leaf stage, suggesting that no new recruitment took place after planting. The apparent synchrony between corn and cutworm phenology in the northern areas of corn production seems more related to the availability and quality of food for young larvae relative to the development of the crop then the time of arrival of moths.

External beam radiation therapy after radical prostatectomy: efficacy and impact on urinary continence.

INTRODUCTION AND OBJECTIVES: The efficacy of adjuvant and salvage external beam radiation (AXRT+SXRT) for prostate cancer after radical prostatectomy (RP) has been debated because of the inability to rule out systemic occult metastasis, uncertainty that radiation eradicates residual local disease and the potential of exacerbating impotency and incontinence. To characterize the effectiveness and treatment morbidity a retrospective review was performed. METHODS: In all, 38 patients received AXRT and 91 received SXRT. The SXRT group was stratified by PSA level, age, race, pathologic stage, margin status, worst Gleason sum, radiation dose and pelvic field. Complications evaluated were impotence and incontinence. Median follow-up was 60.2 months. RESULTS: The 5-y disease-free survival (DFS) rate was 61.3% for AXRT and 36.3% for SXRT. Multivariate analysis of the SXRT cohort showed Gleason score, pathologic stage and pre-XRT PSA to be predictors of disease recurrence. After XRT 26% had worsened continence. CONCLUSIONS: Patients who recur after RP whose pathologic stage is pT2 or pT3c, Gleason score of 8 or higher or pre-XRT PSA is >2.0 ng/dl may have microscopic metastatic disease and a decreased chance of cure with SXRT alone. Continence was further impaired after XRT.

p53 Immunostaining guided laser capture microdissection (p53-LCM) defines the presence of p53 gene mutations in focal regions of primary prostate cancer positive for p53 protein.

OBJECTIVES: A wide range of p53 mutations (5-65%), detected by various methods, has been reported in primary prostate cancers (CaP). IHC staining of radical prostatectomy specimens shows marked heterogeneity of focally distributed p53-positive cells. However, a significant relationship between the focal staining of p53 and cancer recurrence after radical prostatectomy has been noted. Increased frequency of p53 mutations has been generally observed in advanced stage CaP and metastatic prostate cancer cell lines. The significance of focal p53 immunostaining in primary CaP remains uncertain with respect to the p53 gene mutation or tumor progression. The goal of this study was to evaluate p53 gene mutations in focal regions of primary prostate cancers positive by p53 immunostaining. METHODS: Whole-mount prostates from men with clinically organ-confined prostate cancer were immunostained for p53 protein. Laser capture microdissection (LCM) was used to harvest p53 positive cells from areas of tumor and prostatic intraepithelial neoplasia and benign gland. DNA from microdissected cells were amplified for p53 exons 5-8 by polymerase chain reaction (PCR) and analyzed for mutations by single strand conformation polymorphism and DNA sequencing. Mutation analysis of the p53 gene exons 5-8 was performed in the p53 immunostaining positive focal regions (1+ to 4+) of whole-mount prostate sections from 16 patients. RESULTS: Of 16 patients with p53 IHC positive tumors, 11 (69%) had p53 gene mutations as determined by DNA sequence analysis. However, randomly microdissected tumor cells from 4 of 18 patients (22%) negative for p53 IHC also demonstrated mutations in the p53 gene. A significant fraction of prostate tumors with focally positive immunostaining for p53 have been confirmed to contain mutations in the p53 gene. CONCLUSIONS: p53 immunostaining guided LCM combined with DNA-based analyses emphasizes the presence of focal p53 mutations in primary prostate cancers and underscores the significance of previous observations showing a correlation between focal p53 immunostaining in primary CaP and cancer recurrence after radical prostatectomy.

Pentosan polysulfate decreases prostate smooth muscle proliferation and extracellular matrix turnover.

Benign prostatic hyperplasia (BPH) involves proliferation of smooth muscle cells and increased deposition of extracellular matrix (ECM). We recently found that pentosan polysulfate (PPS) has marked effects on growth and ECM of smooth muscle cells derived from vascular tissues. We examined smooth muscle cells cultured from human prostates and the effects of PPS on their growth and ECM production. Fragments of surgical prostatectomy specimens were diced, digested with collagenase (0.01%), and placed in culture medium supplemented with 20% fetal bovine serum. Outgrowths of elongated cells were characterized by light microscopic examination and immunohistochemical techniques by the presence of F-actin, alpha-smooth muscle actin, and myosin, which is a characteristic of smooth muscle cells. Two independent isolates were propagated, and growth curves and ECM production were assessed in the presence and absence of PPS (10 or 100 microg/ml). PPS decreased cell number beginning at day 1 and throughout the incubation period, up to 4 days. The amount of the ECM degradative enzymes, metallo-proteinases MMP-9 and MMP-2, was examined by zymography. PPS did not alter the amount of MMP-2 in the supernatants but MMP-9 was increased 234.4 +/- 17.23-fold over control cells. Tissue inhibitor of MMP (TIMPS), examined by reverse zymography, increased 200% over control. The amount of alpha I type (IV) and alpha I type (I) collagen released in the supernatant, measured by ELISA, significantly decreased in PPS-treated cultures. In conclusion, we found that the administration of PPS decreased proliferation as well as ECM production in prostate smooth muscle. Since smooth muscle proliferation and ECM are involved in the pathophysiology of BPH, PPS may have therapeutic potential.

Primary hormonal therapy for prostate cancer: experience with 135 consecutive PSA-ERA patients from a tertiary care military medical center.

The use of prostate specific antigen (PSA) in the 1990s has brought on a stage migration of prostate cancer. Despite that, many men have still presented with metastatic prostate cancer in the past decade. The use of primary hormone therapy in the PSA era at a tertiary care Army Medical Center is studied in this paper. Charts were reviewed of 135 men who were diagnosed with metastatic prostate cancer and treated with hormone therapy as a primary treatment between 1989 and 1995. Statistical analysis was used to determine significant predictor variables on the time to disease progression. In univariate analysis clinical stage, pretreatment alkaline phosphatase and nadir PSA values were significant predictors of time to progression. Race and type of treatment were not. In multivariate analysis the relative risk of progression was 3.2 for patients with an alkaline phosphatase >252 and 16.5 for patients with a nadir >2.0. This study supports the argument that racial disparities in prostate cancer outcomes are due to access to care. Furthermore, the survival rate for patients with D-2 disease is better than in the pre PSA studies. Clinical stage, pretreatment alkaline phosphatase and PSA nadir can be used to predict response for those men presenting with metastatic prostate cancer.

Androgen deprivation in men with prostate cancer is associated with an increased rate of bone loss.

The objective of this work was to determine the effect of androgen deprivation therapy (ADT) on rates of bone mineral density (BMD) loss in men with prostate cancer. It was a prospective study comparing men receiving ADT to age matched controls for 2 y. Subjects received a history, physical exam, bone mineral density measurement, and laboratory evaluation every 6 months. Thirty-nine subjects receiving continuous ADT for prostate cancer (subjects) were compared to 39 age-matched controls not receiving ADT (controls). Twenty-three subjects and 30 controls completed the study through 24 months. Men in the ADT group demonstrated greater rates of bone mineral density loss than men in the control group at every site except the lumbar spine. Twenty-four month per cent of bone mineral density loss is presented as mean+/-standard error (s.e.). At the distal forearm, the ADT group value was -9.4%+/-1.0% and -4.4%+/-0.3% for controls (P<0.0005). The ADT group femoral neck values were -1.9%+/-0.7% and 0.6%+/-0.5% in the control group (P=0.0016). The ADT group total hip value was -1.5%+/-1.0% and 0.8%+/-0.5% in the control group (P=0.0018). The ADT group trochanter value was -2.0%+/-1.3% and -0.1%+/-0.5% in the control group (P=0.0019). The ADT group lumbar spine value was -0.2%+/-0.8 % and 1.1%+/-0.6% in the control group (P=0.079). Our data demonstrate greater rates of bone mineral density loss in men receiving androgen deprivation therapy for prostate cancer.

A novel human cancer culture model for the study of prostate cancer.

Research into molecular and genetic mechanisms underlying prostate carcinogenesis would be greatly advanced by in vitro models of prostate tumors representing primary tumors. We have successfully established an immortalized human prostate epithelial (HPE) cell culture derived from a primary tumor with telomerase. The actively proliferating early passaged RC-58T cells were transduced through infection with a retrovirus vector expressing the human telomerase catalytic subunit (hTERT). A high level of telomerase was detected in RC-58T/hTERT cells but not RC-58T cells. RC-58T/hTERT cells are currently growing well at passage 50, whereas RC-58T cells senesced at passage 7. RC-58T/hTERT cells exhibit transformed morphology. More importantly, these immortalized cells showed anchorage-independent growth as they formed colonies in soft agar and grew above the agar layer. Expression of androgen-regulated prostate specific gene NKX3.1 and epithelial specific cytokeratin 8 (CK8) but not prostate specific antigen (PSA) and androgen receptor was detected in RC-58T/hTERT cells. Prostate stem cell antigen (PSCA) and p16 were also expressed in this cell line. RC-58T/hTERT cells showed growth inhibition when exposed to retinoic acid and transforming growth factor (TGF)-beta1 known potent inhibitors of prostate epithelial cell growth. A number of chromosome alterations were observed including the loss of chromosomes Y, 3p, 10p, 17p, 18q and the gain of chromosomes 16 and 20. These results demonstrate that this primary tumor-derived HPE cell line retained its transformed phenotypes and should allow studies to elucidate molecular and genetic alterations involved in prostate cancer. This is the first documented case of an established human prostate cancer cell line from a primary tumor of a prostate cancer patient with telomerase.

Improvements in pathologic staging for African-American men undergoing radical retropubic prostatectomy during the prostate specific antigen era: implications for screening a high-risk group for prostate carcinoma.

BACKGROUND: The objective was to compare the changes in pathologic and clinical data over time for African-American (AA) and white men with prostate carcinoma undergoing radical prostatectomy in an attempt to determine the early impact of prostate specific antigen (PSA). METHODS: Data from 195 AA and 587 white men who underwent radical prostatectomy from 1988 to 1999 in an equal access, tertiary, military medical facility were collected. Statistical analysis was used to determine the significance of the changes in the rates of extracapsular extension (ECE), positive margins, pretreatment PSA levels, and age at the time of surgery for each race over time. RESULTS: Comparing 1988-99 results, the authors found that the percentage of AA men with ECE decreased from 100% to 34.8% (P = 0.007), and for white men from 56.9% to 43.2% (P = 0.269). The percentage of AA men with positive margins decreased from 100% to 26.1% (P < 0.0001), and for white men from 41.2% to 27.0% (P = 0.021). Mean age at surgery decreased from 66.6 to 59.9 years for AA men (P < 0.001) and from 65.9 to 61.1 years for white men (P < 0.001). Also, PSA levels decreased from 10.1 to 6.6 ng/dL for white men (P < 0.001) and from 16.5 to 6.5 ng/dL for AA men (P < 0.001). CONCLUSIONS: The authors believe that the decrease in ECE and positive margins in AA men is primarily because of PSA testing, coupled with improved public awareness and equal access to care. It appears reasonable to recommend PSA testing in AA men, who have historically experienced poor outcomes from prostate carcinoma.

Quantitative expression profile of androgen-regulated genes in prostate cancer cells and identification of prostate-specific genes.

Quantitative expression profile of androgen-regulated genes (ARGs) was evaluated in the hormone-responsive prostate cancer cell line LNCaP by serial analysis of gene expression (SAGE). A total of 83,489 SAGE tags representing 23,448 known genes or expressed sequence tags (ESTs) and 1,655 potentially novel sequences have unraveled the transcriptome of LNCaP cells, the most common cell line used in prostate cancer research. Comparison of transcripts between control and R1881-treated LNCaP cells revealed the induction of 136 genes and repression of 215 genes in response to androgen (p < 0.05). Strikingly, a high fraction ( approximately 90%) of ARGs identified in our study has not been described as ARGs previously. A number of prostate-specific transcription factors were among the ARGs identified here. Classification of the ARGs on the basis of biochemical functions revealed that a great majority of ARGs identified in our experimental system appear to be involved in regulation of transcription, splicing, ribosomal biogenesis, mitogenesis, bioenergetics and redox processes. One of the novel aspects of androgen signaling included androgen regulation of genes involved in DNA repair/recombination process. By comparing our LNCaP-C and LNCaP-T SAGE libraries with SAGE tag libraries available at the NCBI-SAGE website, we have identified >200 potential prostate specific/abundant transcripts. The discovery of new prostate-specific genes and ARGs provides a unique opportunity to determine the role of these genes in prostate cell growth, differentiation and tumorigenesis.

Artificial neural network model for the assessment of lymph node spread in patients with clinically localized prostate cancer.

OBJECTIVES: To develop an artificial neural network (ANN) model to predict lymph node (LN) spread in men with clinically localized prostate cancer and to describe a clinically useful method for interpreting the ANN's output scores. METHODS: A simple, feed-forward ANN was trained and validated using clinical and pathologic data from two institutions (n = 6135 and n = 319). The clinical stage, biopsy Gleason sum, and prostate-specific antigen level were the input parameters and the presence or absence of LN spread was the output parameter. Patients with similar ANN outputs were grouped and assumed to be part of a cohort. The prevalence of LN spread for each of these patient cohorts was plotted against the range of ANN outputs to create a risk curve. RESULTS: The area under the receiver operating characteristic curve for the first and second validation data sets was 0.81 and 0.77, respectively. At an ANN output cutoff of 0.3, the sensitivity achieved for each validation set was 63.8% and 44.4%; the specificity was 81.5% and 81.3%; the positive predictive value was 13.6% and 6.5%; and the negative predictive value was 98.0% and 98.1%, respectively. The risk curve showed a nearly linear increase (best fit R(2) = 0.972) in the prevalence of LN spread with increases in raw ANN output. CONCLUSIONS: The ANN's performance on the two validation data sets suggests a role for ANNs in the accurate clinical staging of patients with prostate cancer. The risk curve provides a clinically useful tool that can be used to give patients a realistic assessment of their risk of LN spread.

Introduction to artificial neural networks for physicians: taking the lid off the black box.

BACKGROUND: Over the past 5 years, a steady stream of publications has discussed the use of artificial neural networks (ANNs) for urologic and other medical applications. The pace of this research has increased recently, and deployed products based on this technology are now appearing. Before these tools can be widely accepted by clinicians and researchers, a deeper level of understanding of ANNs is necessary. This article attempts to lay some of the groundwork needed to facilitate this familiarity. METHODS: A short discussion of neural network history is included for background. This is followed by an in-depth discussion of how and why ANNs work. This discussion includes the relationship between ANNs and statistical regression. An investigation of issues associated with neural networks follows, applicable to both general and urologic-specific applications. RESULTS: Neural networks are computer models that have been studied extensively for over 50 years, with prostate cancer applications since 1994. From a biological viewpoint, ANNs are artificial analogues of data structures that exist in nervous systems. From a numeric viewpoint, ANNs are matrices of numbers whose values comprise knowledge that is distilled from historic databases. Many types of neural networks are analogous to well-known statistical methods. CONCLUSIONS: ANNs are complex numeric constructs, but no more complex than similar statistical methods. However, several issues associated with neural network derivation demand that developers apply rigorous engineering practices in their studies.