Reliability of volume or age-adjusted prostate specific antigen to improve diagnostic accuracy

Abstract

Prostate Specific Antigen (PSA), PSA density, PSA excess, and PSA age-specific range have been compared in 331 subjects (35 cancers, 296 benign) with equivocal findings at screening by PSA or palpation + transrectal ultrasonography. The specificity of PSA using a cut-off of 4 ng/ml was only 0.49. A cutoff of 10 ng/ml improved the specificity to 0.88 but the sensitivity decreased from 0.97 to 0.63. The sensitivity and specificity for PSA density (cutoff 0.15), PSA excess (cutoff -5), and age-specific range were 0.86 and 0.80, 0.80 and 0.82, and 0.91 and 0.67, respectively. Avoided benign biopsies vs missed cancers on the basis of PSA density, PSA excess and PSA age-specific range would have been 19 and 5, 26 and 7, and 16 and 3, respectively. Methods adjusting for prostate volume allow a better interpretation of PSA values and may reduce the benign biopsy rate, but the consequent cost of delayed cancer diagnoses when the choice of biopsy is based only on these methods seems unacceptable.