DETECTION OF CIRCULATING PROSTATE DERIVED CELLS IN PATIENTS WITH PROSTATE ADENOCARCINOMA IS AN INDEPENDENT RISK FACTOR FOR TUMOR RECURRENCE ARNAUD MEJEAN, GIOVANNA VONA, BERTRAND NALPAS, DIANE DAMOTTE, NICOLE BROUSSE, YVES CHRETIEN, BERNARD DUFOUR, BERNARD LACOUR, CHRISTIAN BRÉCHOT AND PATRIZIA PATERLINI-BRÉCHOT* From the Service d’Urologie, Service de Biochimie A, and Service d’Anatomie Pathologique, Ho ˆpital Necker, Paris, France ABSTRACT Purpose: To determine whether the presence of prostate-derived cells in the peripheral blood circulation is a marker of prostate cancer and to define the clinical impact of the test. Materials and Methods: We tested the peripheral blood of 99 patients with prostate adenocar- cinoma (PAC), 79 of them undergoing radical prostatectomy, and 92 controls (31 healthy volun- teers, 50 patients with adenoma and 11 with prostatitis) using a highly controlled procedure including reverse-transcriptase polymerase chain reaction (RT-PCR) targeted to prostate-specific antigen (PSA) mRNA. Patients were followed for 26  12 (range: 4 to 49) months. Forty tumor tissues were analyzed by immunohistochemistry for expression of p53 and E-cadherin antigens. Results: Thirty three (33%) patients with PAC and 2 (2%) controls scored positive (p 0.0001) for the test. Detection of circulating prostatic cells was associated with development of metas- tases (p 0.001), with relapse (p 0.001) and with a serum PSA level at diagnosis higher than 15 ng./ml. (p = 0.009). The rate of development of metastases according to time was significantly higher in patients who scored positive for the test (p 0.04). In a multivariate analysis, only the RT-PCR test was an independent risk factor associated with relapse (RR: 6.7). Finally, E-cadherin expression was significantly lower in the tumor tissues of positive patients as compared with those who scored negative for the test (p 0.01). Conclusions: This RT-PCR procedure, performed at diagnosis and with appropriate controls, is a clinically useful assay in evaluating the risk of tumor recurrence after radical prostatectomy in patients with PAC. KEY WORDS: prostatic neoplasms, reverse transcriptase-polymerase chain reaction, prostate-specific antigen, E-cadherin, prognosis The incidence of prostatic adenocarcinoma (PAC), the sec- ond most common malignancy in men, has increased steadily since the 1930s, having the highest incidence rate (from 50 to more than 100 new cases per 100,000) in western countries, the United States and Northern Europe. During this same time frame, the death rate for the disease has doubled. 1, 2 The problem presented by PAC is that currently it is difficult to predict the clinical course of the disease for individuals. Some men survive untreated for many years with localized dis- ease. 3 In contrast, in other men, the tumor metastasizes rapidly, killing the patient within a year from diagnosis. Because of this risk of poor prognosis, extensive screening programs have been suggested for men starting at age 50 years since, if completely localized, prostatic cancer can be cured by surgery alone. 1 Actually, approximately 50% of men with PAC have clini- cally advanced, non organ-confined, disease at the time of diag- nosis. Unfortunately, one third of the remaining 50% of men with organ-confined cancer, as determined by clinical staging, are then found to have invasive disease, as determined by pathological staging performed after prostatectomy. 4 Serum prostate specific antigen (PSA) and histological Gleason score of the biopsy specimen can help in predicting pathologic stage and decrease the incidence of pathologic upstaging. 5 Although this is an improvement, it is clear that significant inaccuracies in preoperative staging remain. New criteria to define the aggressive and metastatic potential of early prostate cancer are needed, 6 particularly in view of recent studies questioning the benefit of radical prostatec- tomy over watchful waiting in early stage PAC. 1 The formation of metastasis involves multisequential events, including, as an early step, tumor cell detachment from primary lesion into the peripheral blood circulation. 2, 7 Therefore, RT-PCR based studies aimed to identify hematog- enous spreading of prostate derived cells have been per- formed in the last few years. 8, 9 A main concern however remains regarding the specificity and the actual clinical im- pact of these RT-PCR based methods. 8, 10, 11 The primary objective of the present work was to test the clinical impact of a highly controlled RT-PCR assay and of a specifically designed study aimed at identifying prostate- derived circulating cells in patients with prostate cancer. We show that a positive test significantly correlates with the risk of developing metastases. In a multivariate analysis, detec- tion of circulating prostate-derived cells was the only risk factor independently associated with tumor recurrence. E-cadherin expression in tumor tissues was also significantly lower in patients who scored positive than in those who scored negative for the assay, supporting the idea that circu- lating prostate-derived cells represent, in these conditions, a marker of prostate cancer. Taken together, our results dem- onstrate that a highly controlled, carefully performed RT- PCR procedure may represent a further noninvasive assay Accepted for publication January 19, 2000. * Requests for reprints: INSERM U 370, Faculte ´ Necker, 75015 Paris, France. Supported by grants from ARC (Association pour la Recherche sur le Cancer), AP (Assistance Publique, contract N°CRC950191), and LNC (Ligue Nationale contre le Cancer). 0022-5347/00/1636-2022/0 THE JOURNAL OF UROLOGY ® Vol. 163, 2022–2029, June 2000 Copyright © 2000 by AMERICAN UROLOGICAL ASSOCIATION,INC. ® Printed in U.S.A. 2022