Head to head comparison of three Prostate Cancer Risk Calculators

Abstract

Background and objectives: Prostate cancer (PCa) risk calculators are potentially useful clinical tools that can help decide if a prostate biopsy (PB) should be performed. Its utilization may avoid some unnecessary biopsies, complications, anxiety and bother. Prostate Biopsy Collaborative Group (PBCG), the Prostate Cancer Research Foundation (SWOP) and Prostate Cancer Prevention Trial (PCPT) are three calculators used worldwide. There are known geographic variations in PCa incidence that are not explained only by local health policies or culture. Our study objective was to compare diagnostic accuracy and validate SWOP, PCPT and PBCG use in the Portuguese population that underwent PB. Patients and methods: This study retrospectively included 179 patients that underwent PB between the 1st of January and 31st of October 2021. Risks of detectable cancer and high risk PCa were calculated using the three before mentioned risk calculators. Three scenarios with different definitions of clinically significant PCa (csPCa) were considered: ISUP ≥3 or more than 2 cores positive for ISUP 2 on the 12-core biopsy (scenario A); ISUP ≥2 (scenario B); ISUP ≥1 (scenario C). Predictive accuracy was assessed by analysis of the area under the receiver-operator characteristic (ROC) curve (AUC). Paired-sample analysis was performed to observe statistically significant differences between AUCs. Results: Patients were 69.13 ± 8.37 years-old, median PSA was 11 ng/mL and malignancy was registered in 77.1%. The SWOP risk calculator was superior to its counterparts with significant differences when compared with PBCG in the first scenario (p=0.012; p=0.017; p=0.032; p=0.048) and PCPT (p=0.033; p=0.044) in the second scenario and third scenario. However, PCPT and PBCG showed to be useful as their AUCs were superior to 0.70 in almost every circumstance. Using SWOP recommended thresholds, 21(11.7%) biopsies would be spared and, in each scenario, respectively, a sensitivity of 93.9%, 96.5% and 93.5%, and a specificity of 29.7%, 33.3% and 31.7%, was observed. Conclusions: This study is the first, to our knowledge, to compare these three calculators in a Portuguese population,. Overall, all calculators showed competence in terms of predictive accuracy. Yet, SWOP showed to be the superior nomogram for the Portuguese population and should be preferred.