ADULT UROLOGY THE ROLE OF TRANSRECTAL ULTRASOUND-GUIDED BIOPSY-BASED STAGING, PREOPERATIVE SERUM PROSTATE-SPECIFIC ANTIGEN, AND BIOPSY GLEASON SCORE IN PREDICTION OF FINAL PATHOLOGIC DIAGNOSIS IN PROSTATE CANCER PERINCHERY NARAYAN, M.D., VISWANATHAN GAJENDRAN, M.D., STEPHEN I? TAYLOR, M.D., ASHUTOSH TEWARI, M.D., JOSEPH C. PRESTI, JR., M.D., RAYMOND LEIDICH, M.D., RICHARD LO, M.D., KEITH PALMER, M.D., KATSUTO SHINOHARA, M.D., ANDJOSEPH T. SPAULDING, M.D. ABSTRACT-Objectives. To evaluate the role of ultrasound-guided systematic and lesion-directed biopsies, biopsy Gleason score, and preoperative serum prostate-specific antigen (PSA) as three objective and repro- ducible variables to provide a reliable combination in preoperative identification of risk of extraprostatic ex- tension in patients with clinically localized prostate cancer. Methods. The case records of 813 patients who underwent radical prostatectomy for clinically localized prostate cancer were analyzed. All had multiple systematic biopsies, two to three from each lobe, in addition to lesion-directed biopsies. Additionally, biopsies were done on seminal vesicles (SVs), if abnormal. Based on biopsy results, patients were classified as having Stage Bl (T2a-T2b) or B2 (T2c) disease, depending on whether biopsies from one or both I’obeswere positive and Stage C (r3) if there was evidence of SV involve- ment by biopsy or biopsies from areas of extracapsular extension as seen on transrectal ultrasound (TRUS) were positive. Logistic regression analyses with log likelihood chi-square test was used to define the correla- tion between individual as well as combination of preoperative variables and pathologic stage. Results. On final pathologic examination, 473 (58%) patients had organ-confined disease, 188 (23%) had extracapsular extension (ECE), with or without positive surgical margins, and 72 (9%) had SV involvement. Eighty (10%) patients had pelvic lymph node metastases. Biopsy-based staging was superior to clinical stag- ing in predicting final pathologic diagnosis. Logistic regression analyses revealed that the combination of biopsy-based stage, preoperative serum PSA, and biopsy Gleason score provided the best prediction of final pathologic stage. Probability plots constructed with these data can provide significant information on risk of extraprostatic extension in individual patients. Conclusions. This study demonstrates that TRUS-guided systematic biopsy in combination with preopera- tive serum PSA and biopsy Gleason score may provide a cost-effective approach for management decisions and prognostication in patients with prostate cancer. UROLCXT 46: 205-2 12, 1995. R adical prostatectomy is most effective for prostate cancer when the disease is organ con- fined.r Among patients with prostate cancer, 60% are thought to have organ-confined disease based on current staging modalities.2,3 However, 40% to 60% of these patients are found to have extrapro- static extension or pelvic lymph node involve- From the Departments of Urology, University of Florida, Gainesville, Florida, University of California School of Medicine, San Francisco, Stanford University, Palo Alto, St. Francis Memorial Hospital, San Francisco, Mt. Diablo Medical Center, Concord, Oakland Naval Hospital, Oakland, and Kaiser Hospital, Santa Rosa, California Perinchery Narayan, M.D., Division of Urology, PO Box 100247, University of Florida, Gainesville, FL 32610-0247 Submitted (Rapid Communication): February 1, 1995, accepted (with revisions): April 6, 199.5 ment in the final pathologic analyses. These pa- tients may not be curable by surgery alone and re- quire alternative or adjunctive management pro- tocols.2-8 The existing techniques for staging are inadequate and need improvement.gy10 Accurate staging would not only allow proper patient se- lection for localized methods of management but would also help in prognosis and appropriate uti- lization of adjunctive therapies. Although digital rectal examination (DRE) is a simple and traditional method of staging, its ma- jor drawbacks are that it underestimates tumor extent in a majority of patients and is subject to interpretational errors, based on the experience of the examiner.11J2 Additionally, with the advent of prostate-specific antigen (PSA), a significant num- ber of Tic cancers (that is, cancers diagnosed on 205