An artificial neural network for five different assay systems of prostate-specific antigen in prostate cancer diagnostics

The use and applicability of machine learning algorithms in predicting the surgical outcome for patients with benign prostatic enlargement. Which model to use?

OBJECTIVES

Artificial intelligence (AI) is increasingly used in medicine, but data on benign prostatic enlargement (BPE) management are lacking. This study aims to test the performance of several machine learning algorithms, in predicting clinical outcomes during BPE surgical management.

METHODS

Clinical data were extracted from a prospectively collected database for 153 men with BPE, treated with transurethral resection (monopolar or bipolar) or vaporization of the prostate. Due to small sample size, we applied a method for increasing our dataset, Synthetic Minority Oversampling Technique (SMOTE). The new dataset created with SMOTE has been expanded by 453 synthetic instances, in addition to the original 153. The WEKA Data Mining Software was used for constructing predictive models, while several appropriate statistical measures, like Correlation coefficient (R), Mean Absolute Error (MAE), Root Mean-Squared Error (RMSE), were calculated with several supervised regression algorithms - techniques (Linear Regression, Multilayer Perceptron, SMOreg, k-Nearest Neighbors, Bagging, M5Rules, M5P - Pruned Model Tree, and Random forest).

RESULTS

The baseline characteristics of patients were extracted, with age, prostate volume, method of operation, baseline Qmax and baseline IPSS being used as independent variables. Using the Random Forest algorithm resulted in values of R, MAE, RMSE that indicate the ability of these models to better predict % Qmax increase. The Random Forest model also demonstrated the best results in R, MAE, RMSE for predicting % IPSS reduction.

CONCLUSIONS

Machine Learning techniques can be used for making predictions regarding clinical outcomes of surgical BPRE management. Wider-scale validation studies are necessary to strengthen our results in choosing the best model.

A systematic review of the applications of Expert Systems (ES) and machine learning (ML) in clinical urology

BACKGROUND

Testing a hypothesis for 'factors-outcome effect' is a common quest, but standard statistical regression analysis tools are rendered ineffective by data contaminated with too many noisy variables. Expert Systems (ES) can provide an alternative methodology in analysing data to identify variables with the highest correlation to the outcome. By applying their effective machine learning (ML) abilities, significant research time and costs can be saved. The study aims to systematically review the applications of ES in urological research and their methodological models for effective multi-variate analysis. Their domains, development and validity will be identified.

METHODS

The PRISMA methodology was applied to formulate an effective method for data gathering and analysis. This study search included seven most relevant information sources: WEB OF SCIENCE, EMBASE, BIOSIS CITATION INDEX, SCOPUS, PUBMED, Google Scholar and MEDLINE. Eligible articles were included if they applied one of the known ML models for a clear urological research question involving multivariate analysis. Only articles with pertinent research methods in ES models were included. The analysed data included the system model, applications, input/output variables, target user, validation, and outcomes. Both ML models and the variable analysis were comparatively reported for each system.

RESULTS

The search identified n = 1087 articles from all databases and n = 712 were eligible for examination against inclusion criteria. A total of 168 systems were finally included and systematically analysed demonstrating a recent increase in uptake of ES in academic urology in particular artificial neural networks with 31 systems. Most of the systems were applied in urological oncology (prostate cancer = 15, bladder cancer = 13) where diagnostic, prognostic and survival predictor markers were investigated. Due to the heterogeneity of models and their statistical tests, a meta-analysis was not feasible.

CONCLUSION

ES utility offers an effective ML potential and their applications in research have demonstrated a valid model for multi-variate analysis. The complexity of their development can challenge their uptake in urological clinics whilst the limitation of the statistical tools in this domain has created a gap for further research studies. Integration of computer scientists in academic units has promoted the use of ES in clinical urological research.

Improvement of Prostate Cancer Diagnosis by Detecting PSA Glycosylation-Specific Changes

New markers based on PSA isoforms have recently been developed to improve prostate cancer (PCa) diagnosis. However, novel approaches are still required to differentiate aggressive from non-aggressive PCa to improve decision making for patients.

PSA glycoforms have been shown to be differentially expressed in PCa. In particular, changes in the extent of core fucosylation and sialylation of PSA N-glycans in PCa patients compared to healthy controls or BPH patients have been reported. The objective of this study was to determine these specific glycan structures in serum PSA to analyze their potential value as markers for discriminating between BPH and PCa of different aggressiveness.

In the present work, we have established two methodologies to analyze the core fucosylation and the sialic acid linkage of PSA N-glycans in serum samples from BPH (29) and PCa (44) patients with different degrees of aggressiveness. We detected a significant decrease in the core fucose and an increase in the α2,3-sialic acid percentage of PSA in high-risk PCa that differentiated BPH and low-risk PCa from high-risk PCa patients. In particular, a cut-off value of 0.86 of the PSA core fucose ratio, could distinguish high-risk PCa patients from BPH with 90% sensitivity and 95% specificity, with an AUC of 0.94. In the case of the α2,3-sialic acid percentage of PSA, the cut-off value of 30% discriminated between high-risk PCa and the group of BPH, low-, and intermediate-risk PCa with a sensitivity and specificity of 85.7% and 95.5%, respectively, with an AUC of 0.97. The latter marker exhibited high performance in differentiating between aggressive and non-aggressive PCa and has the potential for translational application in the clinic.

Current biomarkers for diagnosing of prostate cancer

Prostate cancer (PCa) is mostly detected by prostate-specific antigen (PSA) as one of the most widely used tumor markers. But PSA is limited with its low specificity. The prostate health index (phi) can improve specificity over percent free and total PSA and correlates with aggressive cancer. The urinary PCA3 also shows its utility to detect PCa but its correlation with aggressiveness and the low sensitivity at high values are limitations. While the detection of alterations of the androgen-regulated TMPRSS2 and ETS transcription factor genes in tissue of ˜50% of all PCa patients was one research milestone, the urinary assay should only be used in combination with PCA3. Both US FDA-approved markers phi and PCA3 perform equally.

First Trimester Noninvasive Prenatal Diagnosis: A Computational Intelligence Approach

The objective of this study is to examine the potential value of using machine learning techniques such as artificial neural network (ANN) schemes for the noninvasive estimation, at 11-13 weeks of gestation, the risk for euploidy, trisomy 21 (T21), and other chromosomal aneuploidies (O.C.A.), from suitable sonographic, biochemical markers, and other relevant data. A database(1) (1)The dataset can become available for academic purposes by communicating directly with the authors.

Do prostate cancer risk models improve the predictive accuracy of PSA screening? A meta-analysis

BACKGROUND

Despite the extensive development of risk prediction models to aid patient decision-making on prostate screening, it is unknown whether these models could improve predictive accuracy of PSA testing to detect prostate cancer (PCa). The objective of this study was to perform a systematic review to identify PCa risk models and to assess the model's performance to predict PCa by conducting a meta-analysis.

DESIGN

A systematic literature search of Medline was conducted to identify PCa predictive risk models that used at least two variables, of which one of the variables was prostate-specific antigen (PSA) level. Model performance (discrimination and calibration) was assessed. Prediction models validated in ≥5 study populations and reported area under the curve (AUC) for prediction of any or clinically significant PCa were eligible for meta-analysis. Summary AUC and 95% CIs were calculated using a random-effects model.

RESULTS

The systematic review identified 127 unique PCa prediction models; however, only six models met study criteria for meta-analysis for predicting any PCa: Prostataclass, Finne, Karakiewcz, Prostate Cancer Prevention Trial (PCPT), Chun, and the European Randomized Study of Screening for Prostate Cancer Risk Calculator 3 (ERSPC RC3). Summary AUC estimates show that PCPT does not differ from PSA testing (0.66) despite performing better in studies validating both PSA and PCPT. Predictive accuracy to discriminate PCa increases with Finne (AUC = 0.74), Karakiewcz (AUC = 0.74), Chun (AUC = 0.76) and ERSPC RC3 and Prostataclass have the highest discriminative value (AUC = 0.79), which is equivalent to doubling the sensitivity of PSA testing (44% versus 21%) without loss of specificity. The discriminative accuracy of PCPT to detect clinically significant PCa was AUC = 0.71. Calibration measures of the models were poorly reported.

CONCLUSIONS

Risk prediction models improve the predictive accuracy of PSA testing to detect PCa. Future developments in the use of PCa risk models should evaluate its clinical effectiveness in practice.

Prostate-specific antigen and other serum and urine markers in prostate cancer

Prostate-specific antigen (PSA) is one of the most widely used tumor markers, and strongly correlates with the risk of harboring from prostate cancer (PCa). This risk is visible already several years in advance but PSA has severe limitations for PCa detection with its low specificity and low negative predictive value. There is an urgent need for new biomarkers especially to detect clinically significant and aggressive PCa. From all PSA-based markers, the FDA-approved Prostate Health Index (phi) shows improved specificity over percent free and total PSA. Other serum kallikreins or sarcosine in serum or urine show more diverging data. In urine, the FDA-approved prostate cancer gene 3 (PCA3) has also proven its utility in the detection and management of early PCa. However, some aspects on its correlation with aggressiveness and the low sensitivity at very high values have to be re-examined. The detection of a fusion of the androgen regulated TMPRSS2 gene with the ERG oncogene (from the ETS family), which acts as transcription factor gene, in tissue of ~50% of all PCa patients was one milestone in PCa research. When combining the urinary assays for TMPRSS2:ERG and PCA3, an improved accuracy for PCa detection is visible. PCA3 and phi as the best available PCa biomarkers show an equal performance in direct comparisons.

Artificial neural networks and prostate cancer--tools for diagnosis and management

Artificial neural networks (ANNs) are mathematical models that are based on biological neural networks and are composed of interconnected groups of artificial neurons. ANNs are used to map and predict outcomes in complex relationships between given 'inputs' and sought-after 'outputs' and can also be used find patterns in datasets. In medicine, ANN applications have been used in cancer diagnosis, staging and recurrence prediction since the mid-1990s, when an enormous effort was initiated, especially in prostate cancer detection. Modern ANNs can incorporate new biomarkers and imaging data to improve their predictive power and can offer a number of advantages as clinical decision making tools, such as easy handling of distribution-free input parameters. Most importantly, ANNs consider nonlinear relationships among input data that cannot always be recognized by conventional analyses. In the future, complex medical diagnostic and treatment decisions will be increasingly based on ANNs and other multivariate models.

A comparison of models for predicting sperm retrieval before microdissection testicular sperm extraction in men with nonobstructive azoospermia

PURPOSE

We developed an artificial neural network and nomogram using readily available clinical features to model the chance of identifying sperm with microdissection testicular sperm extraction by readily available preoperative clinical parameters for men with nonobstructive azoospermia.

MATERIALS AND METHODS

We reviewed the records of 1,026 men who underwent microdissection testicular sperm extraction. Patient age, follicle-stimulating hormone level, testicular volume, history of cryptorchidism, Klinefelter syndrome and presence of varicocele were included in the models. For the artificial neural network the data set was divided randomly into a training set (75%) and a test set (25%) with n1/n2 cross validation used to evaluate model accuracy, and then modeled with a neural computational system. In addition, a nomogram with calibration plots was developed to predict sperm retrieval with microdissection testicular sperm extraction. We compared these models to logistic regression.

RESULTS

The ROC area for the neural computational system in the test set was 0.641. The neural network correctly predicted the outcome in 152 of the 256 test set patients (59.4%). The nomogram AUC was 0.59 and adequately calibrated. Multivariable logistic regression demonstrated patient age, history of Klinefelter syndrome and cryptorchidism to be significant predictors of sperm retrieval (p <0.05). However, follicle-stimulating hormone and testicular volume were not significant by internal validation.

CONCLUSIONS

We modeled a combination of well described preoperative clinical parameters to predict sperm retrieval using a neural computational system and nomogram with acceptable predictive values. The generalizability of these findings requires external validation.

Risk stratification in prostate cancer screening

Screening for prostate cancer is a controversial topic within the field of urology. The US Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial did not demonstrate any difference in prostate-cancer-related mortality rates between men screened annually rather than on an 'opportunistic' basis. However, in the world's largest trial to date--the European Randomised Study of Screening for Prostate Cancer--screening every 2-4 years was associated with a 21% reduction in prostate-cancer-related mortality rate after 11 years. Citing the uncertain ratio between potential harm and potential benefit, the US Preventive Services Task Force recently recommended against serum PSA screening. Although this ratio has yet to be elucidated, PSA testing--and early tumour detection--is undoubtedly beneficial for some individuals. Instead of adopting a 'one size fits all' approach, physicians are likely to perform personalized risk assessment to minimize the risk of negative consequences, such as anxiety, unnecessary testing and biopsies, overdiagnosis, and overtreatment. The PSA test needs to be combined with other predictive factors or be used in a more thoughtful way to identify men at risk of symptomatic or life-threatening cancer, without overdiagnosing indolent disease. A risk-adapted approach is needed, whereby PSA testing is tailored to individual risk.

Comparative assessment of urinary prostate cancer antigen 3 and TMPRSS2:ERG gene fusion with the serum [-2]proprostate-specific antigen-based prostate health index for detection of prostate cancer

BACKGROUND

We compared urinary prostate cancer antigen 3 (PCA3), transmembrane protease, serine 2 (TMPRSS2):v-ets erythroblastosis virus E26 oncogene homolog (avian) (ERG) gene fusion (T2:ERG), and the serum [-2]proprostate-specific antigen ([-2]proPSA)-based prostate health index (Phi) for predicting biopsy outcome.

METHODS

Serum samples and first-catch urine samples were collected after digital rectal examination (DRE) from consented outpatients with PSA 0.5-20 μg/L who were scheduled for prostate biopsy. The PCA3 score (PROGENSA PCA3, Hologic Gen-Probe) and T2:ERG score (Hologic Gen-Probe) were determined. Measurements of serum PSA, free PSA, and [-2]proPSA (Beckman Coulter) were performed, and the percentages of free PSA (%fPSA) and Phi ([-2]proPSA/fPSA × √PSA) were determined.

RESULTS

Of 246 enrolled men, prostate cancer (PCa) was diagnosed in 110 (45%) and there was no evidence of malignancy (NEM) in 136 (55%). A first set of biopsies was performed in 136 (55%) of all men, and 110 (45%) had ≥1 repeat biopsies. PCA3, Phi, and T2:ERG differed significantly between men with PCa and NEM, and these markers showed the largest areas under the ROC curve (AUCs) (0.74, 0.68, and 0.63, respectively). PCA3 had the largest AUC of all parameters, albeit not statistically different from Phi. Phi showed somewhat lower specificities than PCA3 at 90% sensitivity. Combination of both markers enhanced diagnostic power with modest AUC gains of 0.01-0.04. Although PCA3 had the highest AUC in the repeat-biopsy cohort, the highest AUC for Phi was observed in DRE-negative patients with PSA in the 2-10 μg/L range.

CONCLUSIONS

PCA3 and Phi were superior to the other evaluated parameters but their combination gave only moderate enhancements in diagnostic accuracy for PCa at first or repeat prostate biopsy.

A Calculator for Prostate Cancer Risk 4 Years After an Initially Negative Screen: Findings from ERSPC Rotterdam

BACKGROUND

Inconclusive test results often occur after prostate-specific antigen (PSA)-based screening for prostate cancer (PCa), leading to uncertainty on whether, how, and when to repeat testing.

OBJECTIVE

To develop and validate a prediction tool for the risk of PCa 4 yr after an initially negative screen.

DESIGN, SETTING, AND PARTICIPANTS

We analyzed data from 15 791 screen-negative men aged 55-70 yr at the initial screening round of the Rotterdam section of the European Randomized Study of Screening for Prostate Cancer.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Follow-up and repeat screening at 4 yr showed either no PCa, low-risk PCa, or potentially high-risk PCa (defined as clinical stage >T2b and/or biopsy Gleason score ≥ 7 and/or PSA ≥ 10.0 ng/ml). A multinomial logistic regression analysis included initial screening data on age, PSA, digital rectal examination (DRE), family history, prostate volume, and having had a previous negative biopsy. The 4-yr risk predictions were validated with additional follow-up data up to 8 yr after initial screening.

RESULTS AND LIMITATIONS

Positive family history and, especially, PSA level predicted PCa, whereas a previous negative biopsy or a large prostate volume reduced the likelihood of future PCa. The risk of having PCa 4 yr after an initially negative screen was 3.6% (interquartile range: 1.0-4.7%). Additional 8-yr follow-up data confirmed these predictions. Although data were based on sextant biopsies and a strict protocol-based biopsy indication, we suggest that men with a low predicted 4-yr risk (eg, ≤ 1.0%) could be rescreened at longer intervals or not at all, depending on competing risks, while men with an elevated 4-yr risk (eg, ≥ 5%) might benefit from immediate retesting. These findings need to be validated externally.

CONCLUSIONS

This 4-yr future risk calculator, based on age, PSA, DRE, family history, prostate volume, and previous biopsy status, may be a promising tool for reducing uncertainty, unnecessary testing, and overdiagnosis of PCa.

External validation of an artificial neural network and two nomograms for prostate cancer detection

Background. Multivariate models are used to increase prostate cancer (PCa) detection rate and to reduce unnecessary biopsies. An external validation of the artificial neural network (ANN) "ProstataClass" (ANN-Charité) was performed with daily routine data. Materials and Methods. The individual ANN predictions were generated with the use of the ANN application for PSA and free PSA assays, which rely on age, tPSA, %fPSA, prostate volume, and DRE (ANN-Charité). Diagnostic validity of tPSA, %fPSA, and the ANN was evaluated by ROC curve analysis and comparisons of observed versus predicted probabilities. Results. Overall, 101 (35.8%) PCa were detected. The areas under the ROC curve (AUCs) were 0.501 for tPSA, 0.669 for %fPSA, 0.694 for ANN-Charité, 0.713 for nomogram I, and 0.742 for nomogram II, showing a significant advantage for nomogram II (P = 0.009) compared with %fPSA while the other model did not differ from %fPSA (P = 0.15 and P = 0.41). All models overestimated the predicted PCa probability. Conclusions. Beside ROC analysis, calibration is an important tool to determine the true value of using a model in clinical practice. The worth of multivariate models is limited when external validations were performed without knowledge of the circumstances of the model's development.

Avoiding pitfalls in applying prediction models, as illustrated by the example of prostate cancer diagnosis

BACKGROUND

The use of different mathematical models to support medical decisions is accompanied by increasing uncertainties when they are applied in practice. Using prostate cancer (PCa) risk models as an example, we recommend requirements for model development and draw attention to possible pitfalls so as to avoid the uncritical use of these models.

CONTENT

We conducted MEDLINE searches for applications of multivariate models supporting the prediction of PCa risk. We critically reviewed the methodological aspects of model development and the biological and analytical variability of the parameters used for model development. In addition, we reviewed the role of prostate biopsy as the gold standard for confirming diagnoses. In addition, we analyzed different methods of model evaluation with respect to their application to different populations. When using models in clinical practice, one must validate the results with a population from the application field. Typical model characteristics (such as discrimination performance and calibration) and methods for assessing the risk of a decision should be used when evaluating a model's output. The choice of a model should be based on these results and on the practicality of its use.

SUMMARY

To avoid possible errors in applying prediction models (the risk of PCa, for example) requires examining the possible pitfalls of the underlying mathematical models in the context of the individual case. The main tools for this purpose are discrimination, calibration, and decision curve analysis.

Between-method differences in prostate-specific antigen assays affect prostate cancer risk prediction by nomograms

BACKGROUND

To date, no published nomogram for prostate cancer (PCa) risk prediction has considered the between-method differences associated with estimating concentrations of prostate-specific antigen (PSA).

METHODS

Total PSA (tPSA) and free PSA were measured in 780 biopsy-referred men with 5 different assays. These data, together with other clinical parameters, were applied to 5 published nomograms that are used for PCa detection. Discrimination and calibration criteria were used to characterize the accuracy of the nomogram models under these conditions.

RESULTS

PCa was found in 455 men (58.3%), and 325 men had no evidence of malignancy. Median tPSA concentrations ranged from 5.5 μg/L to 7.04 μg/L, whereas the median percentage of free PSA ranged from 10.6% to 16.4%. Both the calibration and discrimination of the nomograms varied significantly across different types of PSA assays. Median PCa probabilities, which indicate PCa risk, ranged from 0.59 to 0.76 when different PSA assays were used within the same nomogram. On the other hand, various nomograms produced different PCa probabilities when the same PSA assay was used. Although the ROC curves had comparable areas under the ROC curve, considerable differences were observed among the 5 assays when the sensitivities and specificities at various PCa probability cutoffs were analyzed.

CONCLUSIONS

The accuracy of the PCa probabilities predicted according to different nomograms is limited by the lack of agreement between the different PSA assays. This difference between methods may lead to unacceptable variation in PCa risk prediction. A more cautious application of nomograms is recommended.

Outcome prediction for prostate cancer detection rate with artificial neural network (ANN) in daily routine

BACKGROUND

We evaluated the use of the artificial neural network (ANN) program "ProstataClass" of the Department of Urology and the Institute of Medical Informatics at the Charité-Universitätsmedizin Berlin in daily routine to increase prostate cancer (CaP) detection rate and to reduce unnecessary biopsies.

MATERIALS AND METHODS

From May 2005 to April 2007, a total of 204 patients were included in the study. The Beckman Access PSA assay was used, and pretreatment prostate specific antigen (PSA) was measured prior to digital rectal examination (DRE) and 12 core systematic transrectal ultrasound (TRUS) guided biopsies. The individual ANN predictions were generated with the use of the ANN application for the Beckman Access PSA and free PSA assays, which relies on age, PSA, percent free prostate specific antigen (%fPSA), prostate volume, and DRE. Diagnostic validity of total prostate specific antigen (tPSA), %fPSA, and the ANN was evaluated by ROC curve analysis.

RESULTS

PSA and %fPSA ranged from 4.01 to 9.91 ng/ml (median: 6.65) and 5% to 48% (median: 15%), respectively. Of all men, 46 (22.5%) demonstrated suspicious DRE findings. Total prostate volume ranged from 7.1 to 119.2 cc (median: 35). Overall, 71 (34.8%) CaP were detected. Of men with suspicious DRE, 28 (60.9%) had CaP on initial biopsy. The ANN was 78% accurate in the original report. The AUC of ROC curve analysis was 0.51 for PSA, 0.66 for %PSA, and 0.72 for the ANN-Output, respectively.

CONCLUSIONS

Our results in this independent cohort show that ANN is a very helpful parameter in daily routine to increase the CaP detection rate and reduce unnecessary biopsies.

Internal validation of an artificial neural network for prostate biopsy outcome

OBJECTIVES

To carry out an internal validation of the retrospectively trained artificial neural network (ANN) 'ProstataClass'.

METHODS

A prospectively collected database of 393 patients undergoing 8-12 core prostate biopsy was analyzed. Data of these patients were applied to the online available ANN 'ProstataClass' using the Elecsys total prostate-specific antigen (tPSA) and free PSA (fPSA) assays. Beside the internal validation of the ANN 'ProstataClass' an additional ANN (named as ANN internal validation: ANNiv) only using the 393 prospective patient data was evaluated. The new ANN model was constructed with the MATLAB Neural Network Toolbox. Diagnostic accuracy was evaluated by receiver operator characteristic (ROC) curves comparing the areas under the ROC curves (AUC) and specificities at 90% and 95% sensitivity.

RESULTS

Within a tPSA range of 1.0-22.8 ng/mL, 229 men (58.3%) had prostate cancer (PCa). tPSA, %fPSA and the number of positive digital rectal examinations (DRE) differed significantly from the cohort of patients of the ANN 'ProstataClass', whereas age and prostate volume were comparable. AUCs for tPSA, %fPSA and the ANN 'ProstataClass' were 0.527, 0.726 and 0.747 (P = 0.085 between %fPSA and ANN). The AUC of the ANNiv (0.754) was significantly better compared with %fPSA (P = 0.021), whereas the AUC of two ANN models built on external cohorts (0.726 and 0.729) showed no differences to %fPSA and the other ANN models.

CONCLUSIONS

Significant differences of DRE status and %fPSA medians decrease the power of the 'ProstataClass' ANN in the internal validation cohort. The effect of retrospective data evaluation the 'ProstataClass' cohort and prospective fPSA measurement may be responsible for %fPSA differences. All ANN models built with different PSA and fPSA assays performed equally if applied to the two cohorts.

Reduced serum selenoprotein P concentrations in German prostate cancer patients

Selenium (Se) is essentially needed for the biosynthesis of selenoproteins. Low Se intake causes reduced selenoprotein biosynthesis and constitutes a risk factor for tumorigenesis. Accordingly, some Se supplementation trials have proven effective to reduce prostate cancer risk, especially in poorly supplied individuals. Because Se metabolism is controlled by selenoprotein P (SEPP), we have tested whether circulating SEPP concentrations correlate to prostate cancer stage and grade. A total of 190 men with prostate cancer (n = 90) and "no evidence of malignancy" (NEM; n = 100) histologically confirmed by prostate biopsy were retrospectively analyzed for established tumor markers and for their Se and SEPP status. Prostate specific antigen (PSA), free PSA, total Se, and SEPP concentrations were determined from serum samples and compared with clinicopathologic parameters. The diagnostic performance was analyzed with receiver operating characteristic curves. Median Se and SEPP concentrations differed significantly (P < 0.001) between the groups. Median serum Se concentrations in the 25th to 75th percentile were 95.9 microg/L (82-117.9) in NEM patients and 81.4 microg/L (67.9-98.4) in prostate cancer patients. Corresponding serum SEPP concentrations were 3.4 mg/L (1.9-5.6) in NEM and 2.9 mg/L (1.1-5.5) in prostate cancer patients. The area under the curve (AUC) of a marker combination with age, PSA, and percent free PSA (%fPSA) in combination with the SEPP concentration, yielded the highest diagnostic value (AUC 0.80) compared with the marker combination without SEPP (AUC 0.77) or %fPSA (AUC 0.76). We conclude that decreased SEPP concentration in serum might represent an additional valuable marker for prostate cancer diagnostics.

Techniques and predictive models to improve prostate cancer detection

The use of prostate-specific antigen (PSA) as a screening test remains controversial. There have been several attempts to refine PSA measurements to improve its predictive value. These modifications, including PSA density, PSA kinetics, and the measurement of PSA isoforms, have met with limited success. Therefore, complex statistical and computational models have been created to assess an individual's risk of prostate cancer more accurately. In this review, the authors examined the methods used to modify PSA as well as various predictive models used in prostate cancer detection. They described the mathematical underpinnings of these techniques along with their intrinsic strengths and weaknesses, and they assessed the accuracy of these methods, which have been shown to be better than physicians' judgment at predicting a man's risk of cancer. Without understanding the design and limitations of these methods, they can be applied inappropriately, leading to incorrect conclusions. These models are important components in counseling patients on their risk of prostate cancer and also help in the design of clinical trials by stratifying patients into different risk categories. Thus, it is incumbent on both clinicians and researchers to become familiar with these tools. Cancer 2009;115(13 suppl):3085-99. (c) 2009 American Cancer Society.