Clinical performance of the Prostate Health Index (PHI) for the prediction of prostate cancer in obese men: data from the PROMEtheuS project, a multicentre European prospective study

作者：Abrate Alberto^*; Lazzeri Massimo; Lughezzani Giovanni; Buffi Nicolomaria; Bini Vittorio; Haese Alexander; de la Taille Alexandre; McNicholas Thomas; Palou Redorta Joan; Gadda Giulio M; Lista Giuliana; Kinzikeeva Ella; Fossati Nicola; Larcher Alessandro; Dell'Oglio Paolo; Mistretta Francesco; Freschi Massimo; Guazzoni Giorgio

来源：BJU International, 2015, 115(4): 537-545.

DOI：10.1111/bju.12907

• [-2]proPSA
• obesity
• prostate biopsy
• prostate health index
• prostate specific antigen

Objectives
To test serum prostate-specific antigen (PSA) isoform [-2] proPSA (p2PSA), p2PSA/free PSA (% p2PSA) and Prostate Health Index (PHI) accuracy in predicting prostate cancer in obese men and to test whether PHI is more accurate than PSA in predicting prostate cancer in obese patients.
Patients and Methods
The analysis consisted of a nested case-control study from the pro-PSA Multicentric European Study (PROMEtheuS) project. The study is registered at http://www.controlled-trials.com/ISRCTN04707454. The primary outcome was to test sensitivity, specificity and accuracy (clinical validity) of serum p2PSA, % p2PSA and PHI, in determining prostate cancer at prostate biopsy in obese men [ body mass index (BMI) = 30 kg/m2], compared with total PSA (tPSA), free PSA (fPSA) and fPSA/tPSA ratio (% fPSA). The number of avoidable prostate biopsies (clinical utility) was also assessed. Multivariable logistic regression models were complemented by predictive accuracy analysis and decision-curve analysis.
Results
Of the 965 patients, 383 (39.7%) were normal weight (BMI < 25 kg/m(2)), 440 (45.6%) were overweight (BMI <25-29.9 kg/m(2)) and 142 (14.7%) were obese (BMI >= 30 kg/m(2)). Among obese patients, prostate cancer was found in 65 patients (45.8%), with a higher percentage of Gleason score >= 7 diseases (67.7%). PSA, p2PSA, % p2PSA and PHI were significantly higher, and % fPSA significantly lower in patients with prostate cancer (P < 0.001). In multivariable logistic regression models, PHI significantly increased accuracy of the base multivariable model by 8.8% (P = 0.007). At a PHI threshold of 35.7, 46 (32.4%) biopsies could have been avoided.
Conclusion
In obese patients, PHI is significantly more accurate than current tests in predicting prostate cancer.

• 出版日期2015-4
• 单位Perugia