BERKELEY, CA (UroToday.com) - Currently available predictive models fail to assist clinical decision making in prostate cancer patients who are potential candidates for radical prostatectomy. As such, new biomarkers would be welcome. In a preliminary single-centre study, we found that p2PSA and its derivatives are strong predictors of prostate cancer characteristics at final pathology after radical prostatectomy. However, these initial findings require validation by an independent set of prospective observations, collected on a different set of patients from different centres and by different investigators. We tested the hypothesis that prostate specific antigen (PSA) isoform p2PSA and its derivates, percentage of p2PSA to free PSA (%p2PSA) and the prostate health index (PHI), predict prostate cancer characteristics at final pathology.
We conducted an observational prospective multicentre European study that included 489 consecutive patients treated with radical prostatectomy for prostate cancer. Logistic regression models were fitted to test the predictors of pT3 stage or / and pathological Gleason score ≥ 7 and to determine their predictive accuracy. The base multivariable model included total PSA, digital rectal examination (negative vs. positive), biopsy Gleason score (≤ 6 vs. 7 vs. ≥ 8), and percentage of positive biopsy cores. Decision curve analysis provided an estimate of the net benefit obtained adding p2PSA, %p2PSA, or PHI to the base multivariable model.
Overall, 344 patients (70%) were affected by pT3 disease or pathological Gleason score ≥ 7, while pT3 disease and pathological Gleason score ≥ 7 were present in 126 patients (26%). In the first step of our analyses, we tested the association of index tests (namely, p2PSA, %p2PSA, and PHI) with adverse pathological characteristics. At univariable analyses, p2PSA, %p2PSA, and PHI were significant predictors of pT3 disease or/and pathologic Gleason score ≥7 (all p ≤ 0.001). Moreover, PHI was the most accurate biomarker predicting pT3 disease or/and pathological Gleason Score ≥ 7 (AUC: 0.69 and 0.74, respectively). In the second step of our analyses, we assessed discrimination added by index tests to the base model. The inclusion of PHI significantly increased the predictive accuracy of the base model by 2.3% (p=0.003) for the prediction of pT3 disease or pathological Gleason Score ≥ 7, as well as 2.4% (p=0.01) when considering pT3 disease and pathological Gleason Score ≥ 7. Finally, in the third step of our analyses, we evaluated clinical implications using decision curve analysis. Despite their independent predictor status at multivariable analysis, models including %p2PSA and PHI did not evidently result in a greater net benefit when plotted against various threshold probabilities.
In conclusion, %p2PSA and PHI are significant predictors of unfavorable prostate cancer characteristics at final pathology. However, %p2PSA and PHI do not provide a greater net benefit to clinical decision making.
Written by:
Nicola Fossati, Nicolo Maria Buffi, Alessandro Larcher, Giorgio Guazzoni, and Massimo Lazzeri as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.
Division of Oncology/Unit of Urology, IRCCS Ospedale San Raffaele – Ville Turro, Milan Italy