D W Lin1, M H Gold, S Ransom, W J Ellis, M K Brawer. 1. Department of Urology, University of Washington, Veteran's Affairs Medical Center and Northwest Prostate Institute, Seattle 98116, USA.
Abstract
PURPOSE: Among the new approaches to enhance the performance of prostate specific antigen (PSA) testing in a biopsy population is the use of the free-to-total PSA as well as the transition zone density, which is calculated by dividing the PSA by the transition zone volume. We compare these manipulations of the PSA to PSA alone in a biopsy population. MATERIALS AND METHODS: We evaluated 917 consecutive men who underwent ultrasound guided biopsy for an elevation in serum PSA or abnormality on digital rectal examination. Total PSA was measured using the Tandem-E or Tandem-R method. Prostate gland volume and transition zone were measured with ultrasound and calculated using the prolate ellipsoid formula. RESULTS: In the overall PSA range 276 men had carcinoma (30.0% of the population), while in the PSA 4.0 to 10.0 ng./ml. range 141 of 477 had cancer (29.6%). Receiver operating characteristics analysis and analysis of variance were performed. In the overall PSA series the Tandem total PSA performed as well as any PSA index to predict carcinoma. In the restricted range of total PSA 4.0 to 10.0 ng./ml. total PSA density as well as transition zone density were more predictive than PSA alone. In both PSA ranges the volume of benign glands was significantly larger than in the prostates exhibiting carcinoma. There was no statistically significant difference in outcomes of analyses between different investigators or different sites of investigation (Veterans Affairs versus university based hospitals). CONCLUSIONS: In this biopsy population transition zone PSA density did not add to the information available with total PSA and gland volume. Neither investigator nor site bias contributed to the failure of transition zone PSA density or PSA density to predict prostatic carcinoma.
PURPOSE: Among the new approaches to enhance the performance of prostate specific antigen (PSA) testing in a biopsy population is the use of the free-to-total PSA as well as the transition zone density, which is calculated by dividing the PSA by the transition zone volume. We compare these manipulations of the PSA to PSA alone in a biopsy population. MATERIALS AND METHODS: We evaluated 917 consecutive men who underwent ultrasound guided biopsy for an elevation in serum PSA or abnormality on digital rectal examination. Total PSA was measured using the Tandem-E or Tandem-R method. Prostate gland volume and transition zone were measured with ultrasound and calculated using the prolate ellipsoid formula. RESULTS: In the overall PSA range 276 men had carcinoma (30.0% of the population), while in the PSA 4.0 to 10.0 ng./ml. range 141 of 477 had cancer (29.6%). Receiver operating characteristics analysis and analysis of variance were performed. In the overall PSA series the Tandem total PSA performed as well as any PSA index to predict carcinoma. In the restricted range of total PSA 4.0 to 10.0 ng./ml. total PSA density as well as transition zone density were more predictive than PSA alone. In both PSA ranges the volume of benign glands was significantly larger than in the prostates exhibiting carcinoma. There was no statistically significant difference in outcomes of analyses between different investigators or different sites of investigation (Veterans Affairs versus university based hospitals). CONCLUSIONS: In this biopsy population transition zone PSA density did not add to the information available with total PSA and gland volume. Neither investigator nor site bias contributed to the failure of transition zone PSA density or PSA density to predict prostatic carcinoma.