Diagnostic Performance of a Random Versus Lesion‐Directed Biopsy of the Prostate From Transrectal Ultrasound

OBJECTIVE: The purpose of this study was to compare cancer detection rates of sonographically suspicious lesion-directed biopsies and random biopsies through transverse examination of prostate halves in 518 patients over 5 years. METHODS: From 1998 to 2002, 518 patients were referred for prostate biopsies because of either elevated prostate-specific antigen (PSA) levels or abnormal digital rectal examination findings. On the basis of transverse examination of prostate halves by transrectal ultrasound, we performed lesion-directed biopsy of 3 to 5 cores if morphologically suspicious lesions existed on sonography or random biopsy of 3 cores if no obvious suspicious lesions existed. Biopsy specimens were put into 2 labeled containers. Pathologic results were correlated with random and lesion-directed guided locations. RESULTS: Nine-hundred fifty-nine results were obtained from 439 random and 520 lesion-directed biopsies in 518 patients. Cancer was detected in 207 patients (40.7%). Patients with cancer who had PSA levels of 10.0 ng/mL or greater had higher proportions of bilateral cancer lesions than those with PSA levels of less than 10 ng/mL (P = 0.03). One hundred ten (25.1%) of 439 normal-appearing halves taken by random biopsy were tumor-positive compared with 200 (38.5 %) of 520 biopsies from halves with sonographically suspicious lesions. Regardless of having random or lesion-directed biopsies, patients with PSA levels of 10 ng/mL or greater had higher positive malignancy rates than those with PSA levels of less than 10.0 ng/mL (P < .001). In about 40 patients, a diagnosis was made by random biopsy from halves that were morphologically normal on sonography, not by lesion-directed biopsy from tumor-suspicious contralateral halves. CONCLUSIONS: Cancer detection rates of lesion-directed biopsies are superior to those of random biopsies regardless of PSA level.