The Newer the Better? Comparison of the 1997 and 2001 Partin Tables for Pathologic Stage Prediction of Prostate Cancer in China

Comparison of three versions of Partin tables to predict final pathologic stage in a Chinese cohort: a decision curve analysis

PURPOSE

To compare Partin tables (PTs) 1997, 2001, and 2007 for their clinical applicability in a Chinese cohort based upon a decision curve analysis (DCA).

METHODS

Clinical and pathologic data of 264 consecutive Chinese patients with clinically localized prostate cancer were used. These patients underwent open radical prostatectomy between 2005 and 2011. DCA quantified the net benefit of different PT versions relating to specific threshold probabilities of established capsular penetration (ECP), seminal vesicle involvement (SVI), and lymph node involvement (LNI).

RESULTS

Overall, ECP, SVI, and LNI were recorded in 23.1, 10.2, and 6.1%, respectively. When the threshold probability was below the prevalence for LNI and ECP predictions, the DCA favored the 2007 version versus the 1997 version for SVI.

CONCLUSIONS

DCA indicates that for low threshold probability, decision models are useful to discriminate the performance differences of three PT versions, although net benefit differences were not apparent. For high threshold probability, there may not be an important benefit from the use of PTs and the current analysis cannot translate into meaningful net gains differences.

Machine learning for improved pathological staging of prostate cancer: a performance comparison on a range of classifiers

OBJECTIVES

Prediction of prostate cancer pathological stage is an essential step in a patient's pathway. It determines the treatment that will be applied further. In current practice, urologists use the pathological stage predictions provided in Partin tables to support their decisions. However, Partin tables are based on logistic regression (LR) and built from US data. Our objective is to investigate a range of both predictive methods and of predictive variables for pathological stage prediction and assess them with respect to their predictive quality based on U.K. data.

METHODS AND MATERIAL

The latest version of Partin tables was applied to a large scale British dataset in order to measure their performances by mean of concordance index (c-index). The data was collected by the British Association of Urological Surgeons (BAUS) and gathered records from over 1700 patients treated with prostatectomy in 57 centers across UK. The original methodology was replicated using the BAUS dataset and evaluated using concordance index. In addition, a selection of classifiers, including, among others, LR, artificial neural networks and Bayesian networks (BNs) was applied to the same data and compared with each other using the area under the ROC curve (AUC). Subsets of the data were created in order to observe how classifiers perform with the inclusion of extra variables. Finally a local dataset prepared by the Aberdeen Royal Infirmary was used to study the effect on predictive performance of using different variables.

RESULTS

Partin tables have low predictive quality (c-index=0.602) when applied on UK data for comparison on patients with organ confined and extra prostatic extension conditions, patients at the two most frequently observed pathological stages. The use of replicate lookup tables built from British data shows an improvement in the classification, but the overall predictive quality remains low (c-index=0.610). Comparing a range of classifiers shows that BNs generally outperform other methods. Using the four variables from Partin tables, naive Bayes is the best classifier for the prediction of each class label (AUC=0.662 for OC). When two additional variables are added, the results of LR (0.675), artificial neural networks (0.656) and BN methods (0.679) are overall improved. BNs show higher AUCs than the other methods when the number of variables raises

CONCLUSION

The predictive quality of Partin tables can be described as low to moderate on U.K. data. This means that following the predictions generated by Partin tables, many patients would received an inappropriate treatment, generally associated with a deterioration of their quality of life. In addition to demographic differences between U.K. and the original U.S. population, the methodology and in particular LR present limitations. BN represents a promising alternative to LR from which prostate cancer staging can benefit. Heuristic search for structure learning and the inclusion of more variables are elements that further improve BN models quality.

Serum level of clusterin and its density in men with prostate cancer as novel biomarkers reflecting disease extension

OBJECTIVES

To assess whether the serum level of clusterin and its density could be used as novel biomarkers of prostate cancer.

METHODS

Sera were obtained from 380 patients with prostate cancer and 120 with benign prostatic hyperplasia. Serum clusterin level was measured by a sandwich enzyme immunoassay, and clusterin density, which was determined by dividing the serum clusterin level by the prostate volume, was also calculated. These findings were analyzed with respect to several clinicopathologic factors.

RESULTS

The mean serum level of clusterin in prostate cancer patients was significantly higher than that in the benign prostatic hyperplasia group. Both the serum clusterin level and clusterin density in prostate cancer patients were significantly associated with major prognostic factors other than biopsy Gleason score. Of the 380 prostate cancer patients, 162 underwent radical prostatectomy and pelvic lymphadenectomy, and 104 and 58 were diagnosed as having organ-confined and extraprostatic diseases, respectively. The clusterin density in patients with organ-confined disease was significantly higher than that in patients with extraprostatic disease; however, there was no significant difference in the serum clusterin level between these 2 groups. Furthermore, biochemical recurrence-free survival in patients with elevated clusterin density was significantly lower than that in patients with normal density.

CONCLUSIONS

These findings suggest that serum clusterin level and its density could serve as a useful practical adjuncts to conventional parameters for estimating the extension of prostate cancer.

Novel predictive tools for Irish radical prostatectomy pathological outcomes: development and validation

AIMS

We developed and validated prostate cancer predictive models for Irish patients, allowing individualised predictions of radical prostatectomy pathological outcomes.

METHODS

Retrospective review of the Irish Prostate Cancer Research Consortium database from 2003 to 2008 was performed. Two predictive models were formulated: a replica of the Partin tables (n = 169) and a look-up table based on PSA and biopsy Gleason Score (n = 253). Clinico-pathological parameters were compared to the Partin data set. Internal validation was performed.

RESULTS

In total, 70% of patients were at clinical stage T1c. 5.8% had a PSA less than 4.1 ng/ml, whereas 25% of the Partin patients had a PSA in this range. Maximal predictive accuracy was seen for seminal vesicle invasion (area under the curve = 72%). Prediction of extra-prostatic extension and lymph node involvement was only equivalent to that of a chance phenomenon.

CONCLUSIONS

Our current results do not support the introduction of the formulated predictive models into routine clinical practice.