Vol. 43 (2): 230-238, March – April, 2017

doi: 10.1590/S1677-5538.IBJU.2016.0145


Simon Robinson 1, Marc Laniado 1, Bruce Montgomery 1
1 Frimley Park Foundation Trust, United Kingdom


Background: Prostate-specific antigen densities have limited success in diagnosing prostate cancer. We emphasise the importance of the peripheral zone when considered with its cellular constituents, the “prostatocrit”.

Objective: Using zonal volumes and asymmetry of glandular acini, we generate a peripheral zone acinar volume and density. With the ratio to the whole gland, we can better predict high grade and all grade cancer. We can model the gland into its acinar and stromal elements. This new “prostatocrit” model could offer more accurate nomograms for biopsy.

Materials and Methods: 674 patients underwent TRUS and biopsy. Whole gland and zonal volumes were recorded. We compared ratio and acinar volumes when added to a “clinic” model using traditional PSA density. Univariate logistic regression was used to find significant predictors for all and high grade cancer. Backwards multiple logistic regression was used to generate ROC curves comparing the new model to conventional density and PSA alone.

Outcome and results: Prediction of all grades of prostate cancer: significant variables revealed four significant “prostatocrit” parameters: log peripheral zone acinar density; peripheral zone acinar volume/whole gland acinar volume; peripheral zone acinar density/whole gland volume; peripheral zone acinar density. Acinar model (AUC 0.774), clinic model (AUC 0.745) (P=0.0105).

Prediction of high grade prostate cancer: peripheral zone acinar density (“prostatocrit”) was the only significant density predictor. Acinar model (AUC 0.811), clinic model (AUC 0.769) (P=0.0005).

Conclusion: There is renewed use for ratio and “prostatocrit” density of the peripheral zone in predicting cancer. This outperforms all traditional density measurements.

Keywords: Acinar Cells; Prostatic Neoplasms; Diagnosis; PSA

[Full Text]