Related Experiment Video
Updated: Mar 15, 2026
A Murine Orthotopic Bladder Tumor Model and Tumor Detection System
Published on: January 12, 2017
A nomogram for prediction of prostate cancer on multi-core biopsy using age, serum prostate-specific antigen,
Aim:
To develop and internally validate two nomograms for predicting the probability of overall and clinically-significant prostate cancer on initial biopsy in a Singaporean population.
Methods:
Data were collected from men undergoing initial prostate biopsy at a single center. The indications for biopsy were serum prostate-specific antigen (PSA) ≥4.0 ng/mL or suspicious digital rectal examination (DRE) findings. Men with PSA >30 ng/mL were excluded. Age, PSA, prostate volume (PV) and DRE were predictors included in our logistic regression model and used to construct two nomograms for overall prostate cancer and clinically-significant (Gleason sum ≥7) cancer detection. Predictive accuracies of our nomograms were assessed using area under curve (AUC) of their receiver-operator characteristic curves. Internal validation was performed using the bootstrap method. Our nomograms were compared to a model based on PSA alone using AUC and decision curve analysis (DCA).
Results:
Out of 672 men analyzed, our positive biopsy rate was 26.2% (n = 176), of which 63.6% (n = 112) had clinically significant disease. Age, PSA, PV and DRE status were all independent risk factors for both overall prostate cancer detection as well as clinically-significant cancer detection (all P < 0.05). Our nomogram outperformed serum PSA for both overall and clinically-significant cancer detection (0.736 vs 0.642, P < 0.001 and 0.793 vs 0.696, P < 0.001, respectively). Using DCA, our nomograms had superior net benefit and net reduction in biopsy rate compared to PSA alone.
Conclusions:
Our nomograms have been shown to be superior to PSA alone, on both AUC and DCA. However, it warrants external validation.
