Accuracy of p2PSA and derivatives (%p2PSA and PHI) in predicting prostate cancer in obese men from a multicenter European study

INTRODUCTION & OBJECTIVES: Obesity is becoming highly prevalent all around the world. Its link with prostate cancer (PCa) is controversial, although it seems associated with increased risk of high-grade Gleason Score (GS) at biopsy. Even if PSA accuracy is not affected by BMI, it remains low. In this study, we tested the hypothesis that [-2]proPSA (p2PSA) and derivatives (%p2PSA and PHI) are more accurate than tPSAin predicting PCa in obese men (BMI ≥ 30). MATERIAL & METHODS: The analysis consisted of a nested case-control study from the PRO-psa Multicentric European Study (PROMEtheuS) project (ISRCTN04707454). The primary outcome was to test sensitivity, specificity and accuracy of serum p2PSA, %p2PSA ([(p2PSA pg/ml)/(fPSA ng/ml · 1000)] · 100) and Beckman Coulter PHI ((p2PSA/fPSA) · √PSA), in identifying PCa in obese (BMI ≥ 30) men (clinical validity), and the number of un-necessary biopsies, which could be avoided (clinical utility). Multivariable logistic regression models were complemented by predictive accuracy analysis and decision curve analysis. RESULTS: Over 965 subjects, 383 (39.7%) were normal-weight (BMI < 25), 440 (45.6%) were overweight (BMI 25-29.9) and 142 (14.7%) were obese patients (BMI ≥ 30). Within obese group, PCa was found in 65 subjects (45.8%), 21 with GS 6 (32.3%), 26 with GS 7 (40.0%), and 6, 10 and 2 with GS 8, 9 and 10 (9.2, 15.4 and 3.1%) respectively. PSA, p2PSA, %p2PSA and PHI were significantly higher, and %fPSA significantly lower in patients with PCa (p<0.001). At univariable accuracy analysis, %p2PSA and PHI were the most accurate predictors of PCa at biopsy. In patients with a BMI ≥30, %p2PSA and PHI significantly outperformed tPSA (p<0.001 vs PHI), fPSA (p<0.001) and %fPSA (p=0.008 vs PHI). At 90% sensitivity, the cut-off of %p2PSA and PHI were, respectively, 1.22 and 35.7 with a specificity of 32.5 and 52.3%. At a %p2PSA cut-off of 1.22 a total of 29 (20.4%) biopsies could have been avoided, but 5 (7.7%) cancers would have been missed: 3 with GS 6 (3+3) and 2 cancers with a GS of 7 (3+4). At a PHI cut-off of 35.7 a total of 46 (32.4%) biopsies could have been avoided but 6 (9.2%) cancers would have been missed: 4 with GS 6 (3+3) and 2 cancers with a GS of 7 (3+4). In multivariable logistic regression models, %p2PSA and PHI achieved independent predictor status, significantly increasing the accuracy, by 9.9 (p=0.005) and 8.8% (p=0.007), respectively, of the multivariable base model consisting of patient age, prostate volume, tPSA, fPSA and %fPSA. CONCLUSIONS: In obese patients, %p2PSA and PHI values are significantly and even more accurate than the currently used tests in determining the presence of PCa and could avoid unnecessary biopsies without missing significant PCa.