Percentage of free PSA in black versus white men for detection and stagingof prostate cancer: A prospective multicenter clinical trial

Objectives. In predominately white populations, measurement of the percenta ge of free prostate-specific antigen (%fPSA) has been shown to enhance the specificity of total PSA testing for prostate cancer while maintaining high sensitivity and to aid in prostate cancer staging. This study evaluated wh ether the %fPSA cutoff that maintained a 95% sensitivity in a white populat ion yielded the same sensitivity and specificity in a black population and whether %fPSA was useful in predicting postoperative pathologic features in blacks. Methods. We evaluated 647 white and 79 black men, prospectively enrolled at prostate cancer screening and surgical referral centers. Subjects were 50 to 75 years old with digital rectal examination findings that were not susp icious for prostate cancer and total PSA values between 4.0 and 10.0 ng/mL. All had undergone needle biopsy of the prostate. Hybritech's Tandem total and free PSA assays were used. Results. Ninety-five percent sensitivity was attained with a %fPSA cutoff o f 25% in both races. Use of this cutoff could have avoided unnecessary biop sies in 20% of white and 17% of black subjects (P = 0.69). In receiver oper ating characteristic (ROC] curve analysis, the area under the curve [AUC) f or %fPSA was significantly higher than for total PSA in both blacks (0.76 v ersus 0.56, P < 0.01) and whites (0.70 versus 0.54, P < 0.001). In both rac es, higher %fPSA values indicated a lower risk of cancer and also predicted favorable pathologic features in radical prostatectomy specimens. Conclusions. A 25% fPSA cutoff detected 95% of cancers and reduced unnecess ary biopsies in both races. Higher %fPSA values were associated with favora ble postoperative histopathologic findings in both races. (C) 2000, Elsevie r Science Inc.