Objective: We studied the usefulness of transrectal sonography, prostate-specific antigen levels, and prostate-specific antigen density as indications for directed and random biopsies of the prostate in patients with possible prostatic cancer.
Materials and methods: A total of 141 patients with increased levels of prostate-specific antigen or abnormal findings on digital rectal examination had transrectal sonography of the prostate and determination of prostate-specific antigen density. Through sonographic visualization, all patients had biopsies of possible cancerous lesions and random biopsies of regions of the prostate that appeared normal. Histologic results were correlated with sonographic findings and determinations of prostate-specific antigen levels and prostate-specific antigen density.
Results: Adenocarcinoma was detected in 40 (28%) of the 141 patients. Transrectal sonography showed an abnormality that was determined by directed biopsy to be a carcinoma in 27 (68%) of the 40 patients. Transrectal sonography showed no carcinoma in 13 patients (32%) for whom random biopsy revealed a tumor. The sensitivity of sonography was 68%, and the specificity was 49%. The combination of sonographic findings suggestive of cancer and increased prostate-specific antigen density had a sensitivity of 75% and a specificity of 75%; we calculated a sensitivity of 72% and a specificity of 56% for the combination of sonographic findings suggestive of tumor and increased levels of prostate-specific antigen. Thirty-nine (97%) of 40 patients with cancer had either sonographic findings suggestive of tumor or increased prostate-specific antigen density, and one (3%) had no evidence of tumor on sonography and a normal prostate-specific antigen density.
Conclusion: Directed and random sonographic biopsies of the prostate are indicated in patients with sonographic findings suggestive of tumor and increased prostate-specific antigen density and in patients with abnormal sonographic findings and normal prostate-specific antigen density. Random biopsies are indicated in patients with normal sonographic findings and increased prostate-specific antigen density. In our series, random biopsies were not indicated in 25 of 26 patients with normal sonographic findings and normal prostate-specific antigen density. Further research on the need for random biopsies when there are no sonographic abnormalities and when prostate-specific antigen densities are not elevated is warranted.