OBJECTIVE: To investigate the clinical significance of prostate-specific antigen (PSA)-related markers, including the precursor form of PSA, using the full-range area under the curve of receiver operating characteristics (AUC-ROC), partial AUC-ROC (pAUC-ROC) and multiple logistic regression analyses. METHODS: Participants consisted of 257 consecutive men (PSA range 4.1-20 ng/mL) undergoing transrectal ultrasonography-guided age-adjusted and prostate volume-adjusted multiple-core prostate biopsy at Gunma University Hospital between January 2003 and May 2005. Sensitivity, specificity, AUC-ROC and pAUC-ROC of the ratio of free PSA to total PSA (free/total PSA), PSA density (PSAD) and PSAD adjusted by transition zone volume, the ratio of [-7/-5] precursor forms of PSA (proPSA) to free PSA (pro/free PSA), the ratio of pro to total PSA and the ratio of pro to free/total PSA (pro/f/t ratio) were investigated. Multiple logistic regression analyses were also carried out to investigate the independency of selected clinical parameters. RESULTS: According to pAUC-ROC analyses, pro/free PSA and the pro/f/t ratio were the two best PSA-related parameters in terms of maintaining high sensitivity and avoiding unnecessary biopsy. Multiple regression analyses revealed that not only pro/free PSA, but also age, findings on digital rectal examination and PSAD were independent parameters for predicting biopsy outcomes. CONCLUSION: Pro/free PSA and pro/f/t ratio may be excellent predictive markers for prostate cancer, allowing unnecessary biopsy to be avoided while maintaining high sensitivity at 90% or 95%, in the PSA range 4-20 ng/mL.
OBJECTIVE: To investigate the clinical significance of prostate-specific antigen (PSA)-related markers, including the precursor form of PSA, using the full-range area under the curve of receiver operating characteristics (AUC-ROC), partial AUC-ROC (pAUC-ROC) and multiple logistic regression analyses. METHODS:Participants consisted of 257 consecutive men (PSA range 4.1-20 ng/mL) undergoing transrectal ultrasonography-guided age-adjusted and prostate volume-adjusted multiple-core prostate biopsy at Gunma University Hospital between January 2003 and May 2005. Sensitivity, specificity, AUC-ROC and pAUC-ROC of the ratio of free PSA to total PSA (free/total PSA), PSA density (PSAD) and PSAD adjusted by transition zone volume, the ratio of [-7/-5] precursor forms of PSA (proPSA) to free PSA (pro/free PSA), the ratio of pro to total PSA and the ratio of pro to free/total PSA (pro/f/t ratio) were investigated. Multiple logistic regression analyses were also carried out to investigate the independency of selected clinical parameters. RESULTS: According to pAUC-ROC analyses, pro/free PSA and the pro/f/t ratio were the two best PSA-related parameters in terms of maintaining high sensitivity and avoiding unnecessary biopsy. Multiple regression analyses revealed that not only pro/free PSA, but also age, findings on digital rectal examination and PSAD were independent parameters for predicting biopsy outcomes. CONCLUSION: Pro/free PSA and pro/f/t ratio may be excellent predictive markers for prostate cancer, allowing unnecessary biopsy to be avoided while maintaining high sensitivity at 90% or 95%, in the PSA range 4-20 ng/mL.
Authors: Lori J Sokoll; Martin G Sanda; Ziding Feng; Jacob Kagan; Isaac A Mizrahi; Dennis L Broyles; Alan W Partin; Sudhir Srivastava; Ian M Thompson; John T Wei; Zhen Zhang; Daniel W Chan Journal: Cancer Epidemiol Biomarkers Prev Date: 2010-05 Impact factor: 4.254