Prostate cancer detection in the "grey area" of prostate-specific antigen below 10 ng/ml: head-to-head comparison of the updated PCPT calculator and Chun's nomogram, two risk estimators incorporating prostate cancer antigen 3

Background: Prostate cancer antigen 3 (PCA3) holds promise in diagnosing prostate cancer (PCa), but no consensus has been reached on its clinical use. Multivariable predictive models have shown increased accuracy over individual risk factors.

Objective: To compare the performance of the two available risk estimators incorporating PCA3 in the detection of PCa in the "grey area" of prostate-specific antigen (PSA) <10 ng/ml: the updated Prostate Cancer Prevention Trial (PCPT) calculator and Chun's nomogram.

Design, setting, and participants: Two hundred eighteen patients presenting with an abnormal PSA (excluding those with PSA >10 ng/ml) and/or abnormal digital rectal examination were prospectively enrolled in a multicentre Italian study between October 2008 and October 2009. All patients underwent ≥12-core prostate biopsy.

Measurements: PCA3 scores were assessed using the Progensa assay (Gen-Probe, San Diego, CA, USA). Comparisons between the two models were performed using tests of accuracy (area under the receiver operating characteristic curve [AUC-ROC]), calibration plots, and decision curve analysis. Biopsy predictors were identified by univariable and multivariable logistic regression. In addition, performance of PCA3 was analysed through AUC-ROC and predictive values.

Results and limitations: PCa was detected in 73 patients (33.5%). Among predictors included in the models, only PCA3, PSA, and prostate volume retained significant predictive value. AUC-ROC was higher for the updated PCPT calculator compared to Chun's nomogram (79.6% vs 71.5%; p=0.043); however, Chun's nomogram displayed better overall calibration and a higher net benefit on decision curve analysis. Using a probability threshold of 25%, no high-grade cancers would be missed; the PCPT calculator would save 11% of biopsies, missing no cancer, whereas Chun's nomogram would save 22% of avoidable biopsies, although missing 4.1% non-high-grade cancers. The small number of patients may account for the lack of statistical significance in the predictive value of individual variables or model comparison.

Conclusions: Both Chun's nomogram and the PCPT calculator, by incorporating PCA3, can assist in the decision to biopsy by assignment of an individual risk of PCa, specifically in the PSA levels <10 ng/ml.