Purpose: We evaluate the diagnostic use of total, free and complexed serum prostate specific antigen (PSA), and their ratios for enhancing the specificity in detecting prostate cancer.
Materials and methods: A total of 354 nonconsecutive men undergoing prostate biopsy were eligible for this retrospective and prospective study. Cancer was found in 122 of these 354 men (34%). Receiver operating characteristics curve analyses were used to calculate and compare the performance of total PSA (Hybritech, San Diego California and Bayer, Tarrytown, New York), complexed PSA (Bayer), percent complexed PSA and percent free PSA. In addition, sensitivity and specificity were calculated and compared.
Results: The area under the receiver operating characteristics curve was highest for percent free PSA, followed by percent complexed PSA, complexed PSA and the 2 total PSA assays (Hybritech and Bayer). The cutoff value of 3.45 ng./ml. for complexed PSA detected the same number of cancers and resulted in 1 additional false-positive case compared with a Hybritech total PSA threshold of 4.0 ng./ml. At sensitivities of 80% to 95%, there were no significant differences for detection comparing the corresponding specificities between Hybritech total PSA and complexed PSA for all 354 men. Complexed PSA alone did not enhance the overall diagnostic accuracy compared with percent free PSA in the Hybritech total PSA range between 4.01 and 6.00 ng./ml., between 6.01 and 10.00 ng./ml., and between 2.50 and 6.00 ng./ml. At sensitivities of 80% to 95% specificity of percent complexed PSA was almost identical to that of percent free PSA except for the Hybritech total PSA range less than or equal to 4.00 ng./ml.
Conclusions: This study suggests complexed PSA is equivalent to total PSA for the early detection of prostate cancer. Percent free PSA outperforms complexed PSA and percent complexed PSA performed equivalently to percent free PSA in all total PSA ranges analyzed between 2.5 and 10 ng./ml.