Partin Tables cannot accurately predict the pathological stage at radical prostatectomy

doi:10.1016/j.ejso.2008.07.013

European Journal of Surgical Oncology (EJSO)

Volume 35, Issue 2, February 2009, Pages 123-128

https://doi.org/10.1016/j.ejso.2008.07.013 Get rights and content

Abstract

Purpose

The Partin Tables represent the most commonly used staging tool for radical prostatectomy (RP) candidates. The Partin Tables' predictions are used to guide the type (nerve preserving RP) and/or the extent (RP with wide resection) of RP. We examined the ability of the Partin Tables' predictions incorrectly assigning the stage at RP.

Methods

The testing of the Partin Tables (external validation) was based on 3105 patients treated with RP at a single European institution. Standard validation metrics were used (area under the receiver operating characteristics curve, AUC) to test the three endpoints predicted by the Partin Tables, namely the presence of extracapsular extension (ECE), seminal vesicle invasion (SVI), and lymph node invasion (LNI).

Results

Ideal predictions are denoted with 100% accuracy vs. 50% for entirely random predictions. For the 2001 version of the Tables the accuracy defined by the AUC was 79.7, 77.8, and 73.0 for ECE, SVI, and LNI, respectively. For the 2007 version of the Tables the corresponding accuracy estimates were 79.8, 80.5, and 76.2. The relationship between predicted probabilities and observed rates was poor.

Conclusion

The Partin Tables are meant to guide clinicians about the safety of nerve bundle preservation at RP, about the need for seminal vesicle resection or for lymphadenectomy. Therefore, the use of the Partin Tables predictions may significantly affect the type and/or the extent of RP. In their present format the Partin Tables are not accurate enough to influence the pre-operative decision making regarding the type or extent of RP.

Introduction

Several nomograms and other statistical models were recently devised to predict pathological stage prior to definitive therapy.1, 2, 3, 4, 5, 6, 7, 8, 9, 10 These tools offer the possibility of using their predictions for the most evidence-based and individualized selection of available treatment alternatives. Among different models, the Partin Tables represent one of the most fundamental stepping stones that led to the development of the field of prognostics in urologic oncology.11, 12, 13, 14 The Partin Tables rely on pretreatment prostate specific antigen (PSA), clinical stage, and biopsy Gleason score to predict the probability of extracapsular extension (ECE), seminal vesicle invasion (SVI) and lymph node invasion (LNI). The Partin Tables help urologists, radiation oncologists, medical oncologists and general practitioners to predict what stage of the disease might be expected, if the prostate were surgically removed. The Table's predictions represent the most widely used guide towards the selection for definitive therapy in Europe and North America. For example, the Partin Tables' predictions of ECE guide urological surgeons with respect to the safety of neurovascular bundle preservation. Patients at low risk of ECE may be considered for neurovascular bundle sparing-radical prostatectomy.15, 16 Conversely, those at high risk should not. The Partin Tables' predictions of SVI can help identifying those at low risk of SVI, in whom the seminal vesicles may be spared. This may improve post-operative erectile and urinary function.⁴ Finally, the Partin Tables are widely used to identify patients at risk of LNI who should be staged with a pelvic lymphadenectomy.17, 18

The original Partin Tables were developed in 1993 and were subsequently updated in 1997.11, 12 The 1997 Tables have become the most widely used staging tool in urologic oncology. Due to prostate cancer stage migration, the Tables were updated in 2001 (n = 5079).13, 14 The validity of the 2001 Tables was confirmed in several populations from North America,¹⁹ Canada,²⁰ and Europe.²¹ The discrimination and calibration properties were compared to the 1997 Tables and showed equally good results.²² The ongoing stage migration prompted Partin and associates to update their series again and the authors included men treated with radical prostatectomy (RP) between 2000 and 2005, which resulted in the 2007 update of the Tables (n = 5730).¹⁴ The updated 2007 version of the Tables represents an important contribution for the urologic community in North America. However, the rate and the type of stage migration that occurred in Europe during the same time period (2000–2005) did not necessarily perfectly replicate the stage and grade migration that occurred in North America. Recent reports suggest that important differences in stage, grade and PSA at presentation may exist between men in the United States and those in Europe.² These differences were shown to result in equally important differences in the pathological stage at RP.2, 23 These differences might undermine the performance characteristics and/or the accuracy of the Partin Tables in European patients, especially since the Partin Tables were developed in men from the United States. Important differences in patient's characteristics may therefore lead to inaccurate predictions. Inaccurate predictions may in turn mislead the clinicians with respect to the safety of neurovascular bundle preservation or the need for seminal vesicle resection or lymphadenectomy. Therefore, prior to the clinical implementation of the 2007 Partin Tables, we decided to assess their accuracy and performance characteristics in patients treated with RP at one tertiary academic centre in Europe.

Section snippets

Patients

To comply with the temporal characteristics of the original Partin Tables' cohort, we relied on 3288 consecutive patients treated with radical prostatectomy for localized prostate cancer at the University of Eppendorf in Hamburg, Germany between January 2000 and December 2005. Of those, 61 were excluded due to missing PSA values, 16 for missing clinical stage information, and 87 for missing biopsy Gleason scores. Men with clinical stages T1a (n = 6), T1b (n = 3), and T3 (n = 10) were also excluded to

Results

Clinical and pathological characteristics of the current study cohort and those of the Partin 2001 and 2007 cohorts are displayed in Table 1. For all three cohorts the data are strikingly similar with respect to clinical stage distribution. Serum PSA distribution of the German cohort and of the 2001 Partin Table cohorts are equally comparable. Conversely, relative to the German cohort, a higher proportion of low serum PSA values were reported in the 2007 Partin cohort. The German cohort

Discussion

The pathologic stage of the disease represents the determinant of treatment choice. For example, patients with a high likelihood of extra prostatic disease (ECE, SVI, or LNI) are not good candidates for observation (active surveillance) or for less definitive treatment modalities, such as brachytherapy or high intensity focused ultrasound.25, 26, 27, 28 The consequence of incorrect staging leads to inappropriate preservation of the neurovascular bundle in patients with ECE or of lack of staging

Conclusion

The Partin Tables guide clinicians with respect to the safety of nerve bundle preservation at RP or with respect to the need for seminal vesicle resection or lymphadenectomy. The use of Partin Tables predictions may significantly affect the type and/or the extent of the surgery that is performed. In their present format the Partin Tables are not accurate enough to warrant their use in pre-operative decision making.

Conflict of interest

The authors have no conflict of interest.

Acknowledgments

Pierre I. Karakiewicz is partially supported by the University of Montreal Health Center Urology Associated, Fonds de la Recherche en Santè du Quebec, the University of Montreal Department Of Surgery and the University of Montreal Health Center (CHUM) Foundation.

Dr. François Péloquin and Dr. Daniel Pharand both contributed to the critical revision of this manuscript.

References (29)

T. Steuber et al.
Validation of a nomogram for prediction of side specific extracapsular extension at radical prostatectomy
J Urol
(2006)
A. Gallina et al.
Development and split-sample validation of a nomogram predicting the probability of seminal vesicle invasion at radical prostatectomy
Eur Urol
(2007)
F.K. Chun et al.
Development and internal validation of a nomogram predicting the probability of prostate cancer Gleason sum upgrading between biopsy and radical prostatectomy pathology
Eur Urol
(2006)
U. Capitanio et al.
Radical prostatectomy for incidental (stage t1a–t1b) prostate cancer: analysis of predictors for residual disease and biochemical recurrence
Eur Urol
(2008)
A.W. Partin et al.
The use of prostate specific antigen, clinical stage and Gleason score to predict pathological stage in men with localized prostate cancer
J Urol
(1993)
A.W. Partin et al.
Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium
Urology
(2001)
D.V. Makarov et al.
Updated nomogram to predict pathologic stage of prostate cancer given prostate-specific antigen level, clinical stage, and biopsy Gleason score (Partin tables) based on cases from 2000 to 2005
Urology
(2007)
M.W. Kattan
Re: Updated nomogram to predict pathologic stage of prostate cancer given prostate-specific antigen level, clinical stage, and biopsy Gleason score (Partin Tables) based on cases from 2000 to 2005
Eur Urol
(2007)
T. Steuber et al.
Transition zone cancers undermine the predictive accuracy of Partin table stage predictions
J Urol
(2005)
H. Augustin et al.
Comparison of accuracy between the Partin tables of 1997 and 2001 to predict final pathological stage in clinically localized prostate cancer
J Urol
(2004)

S. Conrad et al.

Prospective validation of an algorithm with systematic sextant biopsy to predict pelvic lymph node metastasis in patients with clinically localized prostatic carcinoma

J Urol

(2002)

A. Heidenreich et al.

EAU guidelines on prostate cancer

Eur Urol

(2008)

R.C. van den Bergh et al.

Prospective validation of active surveillance in prostate cancer: the PRIAS study

Eur Urol

(2007)

F.K. Chun et al.

Initial biopsy outcome prediction-head-to-head comparison of a logistic regression-based nomogram versus artificial neural network

Eur Urol

(2007)

Cited by (29)

Preoperative Risk Assessment
2016, Prostate Cancer: Science and Clinical Practice: Second Edition
Once a patient is diagnosed with prostate cancer, the decision to proceed with therapy and which therapy is an extremely complex one. Variables that factor into treatment decisions are inherent to the patient (comorbidity, life expectancy, preferences) and to their cancer (disease aggressiveness and potential for cure). Although most decisions are likely made based on physician and patient experiences and bias, objective predictive models based on pretreatment variables allow for more accurate predictions of prognosis and outcome. Pretreatment tools have been used with a variety of different goals including predicting pathologic outcomes, nonprostate-cancer related mortality, and biochemical and prostate-cancer-specific survival. To this end, there now exists an extensive list of predictive tools that can be used as decision aids after an initial diagnosis of prostate cancer. The focus of this chapter will be the most commonly used and studied pretreatment decision tools.
Artificial neural network for predicting pathological stage of clinically localized prostate cancer in a Taiwanese population
2014, Journal of the Chinese Medical Association
We developed an artificial neural network (ANN) model to predict prostate cancer pathological staging in patients prior to when they received radical prostatectomy as this is more effective than logistic regression (LR), or combined use of age, prostate-specific antigen (PSA), body mass index (BMI), digital rectal examination (DRE), trans-rectal ultrasound (TRUS), biopsy Gleason sum, and primary biopsy Gleason grade.
Our study evaluated 299 patients undergoing retro-pubic radical prostatectomy or robotic-assisted laparoscopic radical prostatectomy surgical procedures with pelvic lymph node dissection. The results were intended to predict the pathological stage of prostate cancer (T2 or T3) after radical surgery. The predictive ability of ANN was compared with LR and validation of the 2007 Partin Tables was estimated by the areas under the receiving operating characteristic curve (AUCs).
Of the 299 patients we evaluated, 109 (36.45%) displayed prostate cancer with extra-capsular extension (ECE), and 190 (63.55%) displayed organ-confined disease (OCD). LR analysis showed that only PSA and BMI were statistically significant predictors of prostate cancer with capsule invasion. Overall, ANN outperformed LR significantly (0.795 ± 0.023 versus 0.746 ± 0.025, p = 0.016). Validation using the current Partin Tables for the participants of our study was assessed, and the predictive capacity of AUC for OCD was 0.695.
ANN was superior to LR at predicting OCD in prostate cancer. Compared with the validation of current Partin Tables for the Taiwanese population, the ANN model resulted in larger AUCs and more accurate prediction of the pathologic stage of prostate cancer.
Fuzzy expert system for predicting pathological stage of prostate cancer
2013, Expert Systems with Applications
Citation Excerpt :
Augustin, Sun, Isbarn, Pummer, and Karakiewicz (2012) compared the three versions and did not identify the best one. The Partin tables represent the most widely used guide towards the selection of definitive therapies in Europe and North America (Bhojani et al., 2009). The European Association of Urology (EAU) guidelines recommend the use of Partin tables and Kattan nomogram for prostate cancer staging (Briganti, Karakiewicz, Joniau, & Van Poppel, 2009).
Prostate cancer is the second most common cancer among men, responsible for the loss of half a million lives each year worldwide, according to the World Health Organization. In prostate cancer, definitive therapy such as radical prostatectomy, is more effective when the cancer is organ-confined. The aim of this study is to investigate the performance of some fuzzy expert systems in the classification of patients with confined or non-confined cancer. To deal with the intrinsic uncertainty about the variables utilized to predict cancer stage, the developed approach is based on Fuzzy Set Theory. A fuzzy expert system was developed with the fuzzy rules and membership functions tuned by a genetic algorithm. As a result, the utilized approach reached better precision taking into account some correlated studies.
Development and validation of a UK-specific prostate cancer staging predictive model: UK prostate cancer tables
2012, British Journal of Medical and Surgical Urology
To construct new prostate cancer staging lookup tables based on a dataset collated by the British Association of Urological Surgeons (BAUS) and to validate them and compare their predictive power with Partin tables.
Complete data on 1701 patients was collated between 1999 and 2008 across 57 UK centres. Lookup tables were created for prediction of pathological stage (PS) using PSA level, biopsy Gleason score (GS) and clinical stage, replicating Partin's original approach.
Tables were generated using logistic regression (LR) and bootstrap resampling methods and were internally validated and externally validated using concordance indices (CI) and area under the receiver operating characteristic curve (AUROC) respectively.
The CI and AUROC analyses indicate that Partin tables performed poorly on UK data in comparison with US data.
The UK prostate cancer tables performed better than Partin tables but the predictive power of all models was relatively poor.
The study shows that the predictive power of Partin tables is reduced when applied to the UK population.
Models generated using LR methodology have fundamental limitations, and we suggest alternative modelling methods such as Bayesian networks.
Decision curve analysis to compare 3 versions of Partin Tables to predict final pathologic stage
2012, Urologic Oncology: Seminars and Original Investigations
Citation Excerpt :
However, recently certain limitations were recorded in European patient cohorts. For example, in men from Germany, France, and Italy, the 2007 version of the Partin Tables showed suboptimal performance characteristics [8,9]. This observation prompted us to examine the 1997 and 2001 versions of the Partin Tables with the intent of comparing them with the 2007 version.
To perform a decision curve analysis (DCA) to compare the Partin Tables 1997, 2001, and 2007 for their clinical applicability.
Clinical and pathologic data of 687 consecutive patients treated with open radical prostatectomy for clinically localized prostate cancer between 2003 and 2008 at a single institution were used. DCA quantified the net benefit relating to specific threshold probabilities of extraprostatic extension (EPE), seminal vesicle involvement (SVI), and lymph node involvement (LNI).
Overall, EPE, SVI, and LNI were recorded in 17.8, 6.0, and 1.2%, respectively. For EPE predictions, the DCA favored the 2007 version vs. 1997 for SVI vs. none of the versions for LNI.
DCA indicate that for very low prevalence conditions such as LNI (1.2%), decision models are not useful. For low prevalence rates such as SVI, the use of different versions of the Partin Tables does not translate into meaningful net gains differences. Finally, for intermediate prevalence conditions such as EPE (18%), despite apparent performance differences, the net benefit differences were also marginal. In consequence, the current analysis could not confirm an important benefit from the use of the Partin Tables and it could not identify a clearly better version of any of the 3 available iterations.
Machine learning for improved pathological staging of prostate cancer: A performance comparison on a range of classifiers
2012, Artificial Intelligence in Medicine
Citation Excerpt :
Using the British data and the approach described in [8], we build new lookup tables and assess the methodology itself. The results are compared against the previous validation studies [9–19] and provide additional understanding on Partin tables performances. Second, we propose alternative classifying techniques to build lookup tables for prostate cancer staging.
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 UK data.
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.
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
The predictive quality of Partin tables can be described as low to moderate on UK 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 UK and the original US 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.

View all citing articles on Scopus

¹: Equal contribution.

View full text