Background: This was a pilot study on EUS for locoregional evaluation of prostate cancer.
Objective: Our purpose was to evaluate radial and linear-array EUS in locoregional prostate cancer staging.
Design, setting, patients: From April to December 2005, 23 patients were referred to the Department of Urology with a confirmed or highly suspected diagnosis of prostate cancer on the basis of cytohistologic examination of fragments obtained by transrectal US-guided biopsy or transuretheral means. After institutional review board approval, informed consent was obtained from all patients.
Intervention: An endosonographer and a radiologist with expertise in prostate imaging performed radial and linear EUS examinations without knowledge of the stage of prostate cancer of the referred patients.
Main outcome measurements: Systematic prostatic evaluation by EUS. All patients underwent prostatectomy, and the surgical specimens were analyzed and correlated with EUS findings.
Results: Mean age was 65.91 years, and the mean prostate-specific antigen level was 27.73 ng/mL. Histopathologic study of the surgical specimen revealed adenocarcinoma in 20 of 23, atypical adenomatous hyperplasia in 2 of 23, and sclerosing adenosis in 1 of 23. Staging by EUS for T stage showed different sensitivity (S), specificity (E), and accuracy (A) according to the degree of tumor invasiveness as follows: T1 (S: 51.3%, E: 53.2%, A: 49.1%); T2 (S: 100%, E: 91.67%, A: 95%); T3 (S: 100%, E: 100%, A: 100%). In 3 (3/23) patients EUS did not find a defined lesion, but the surgical specimen showed T1 stage cancer. EUS staging for N stage showed 62.5% sensitivity, 58.33% specificity, and 60% accuracy for N0. Regarding N1, 58.3% sensitivity, 52.50% specificity, and 60% accuracy were found.
Limitations: Uncontrolled, nonrandomized study.
Conclusions: EUS presented high sensitivity, specificity, and accuracy for prostate cancer staging.