DETECTION OF DISSEMINATED PROSTATE CELLS BY REVERSE TRANSCRIPTION-POLYMERASE CHAIN REACTION (RT-PCR): TECHNICAL AND CLINICAL ASPECTS Eva COREY* and Michael J. COREY Genitourinary Research Cancer Laboratory of the Urology Department, the School of Medicine of the University of Washington, Seattle, WA, USA Metastasis of prostate cancer (CaP) to bone is responsible for much of the morbidity and mortality associated with this disease. While the primary goal of current methods of treatment is to prevent, slow down or reverse metastatic spread, the sensitivity of today’s conventional staging techniques is insufficient to detect micrometastatic disease. Radical prostatectomy is intended to be a complete cure, but 15–30% of these cases recur, because CaP cells that have escaped the prostate prior to surgery give rise to metastases. Although macroscopic techniques such as the bone scan cannot detect micrometastases, new molecular approaches that are under development may prove to be capable of filling this need. There is now hope that detection of prostate cells in circulation may provide a new and highly sensitive staging tool. Because prostate-specific antigen (PSA) is produced almost exclu- sively by prostate epithelial cells (Clements and Mukhtar, 1994; Levesque et al., 1995a,b; Oesterling, 1991; Zarghami and Diaman- dis, 1996; Zarghami et al., 1997), the presence of PSA-positive cells in circulation clearly indicates an abnormal condition, and a strong correlation between the appearance of prostate cells in circulation and the advent of metastasis might be expected. Current models of metastasis hold that cells ‘‘shed’’by primary tumors are borne by the circulation (blood and lymphatics) to other organs, where they are deposited and form secondary tumors. It is not known whether these circulating cells have metastatic potential, or whether they must undergo further changes to cause metastases. Protocols based on the reverse transcription-polymerase chain reaction (RT-PCR) have been developed in response to the need for identification and widespread adoption of superior methods for detecting these extra-prostatic cells, so that more systematic studies with long-term follow-up may be undertaken. Currently, the mRNAs encoding PSA, prostate-specific membrane antigen (PSMA), and human glandular kallikrein (hK2) are the 3 messages most frequently used for detection of prostate cells, but these messages, while prostate-specific, are not cancer-specific, and therefore yield little or no information about the metastatic potential of the cells detected. Markers related to cancer progres- sion are desired from the viewpoints of both prognostic methods and basic understanding. The limit of detection (LoD) of RT-PCR is better than that of other methods in current use, such as flow cytometry and immuno- histochemistry (Pelkey et al., 1996). There is therefore consider- able interest in the use of RT-PCR for detection and possible quantification of prostate cells in peripheral blood (PB), bone marrow (BM) and lymph node (LN) samples from CaP patients. However, despite the promise of this method, it is not yet clear whether RT-PCR will prove to be of clinical value in prostate- cancer staging. In this review we will summarize published reports on RT-PCR detection of circulating prostate cells. Because the field is develop- ing rapidly, there have been several significant advances since the last major reviews were published (Gomella et al., 1997; Olsson et al., 1997), including the use of longer primers with 2-temperature PCR for specificity, development of internal controls and targeting of the hK2 message. There have also been a number of additional reports describing RT-PCR protocols and clinical results in CaP. With a very few exceptions, we have restricted the scope of this review to published articles. The review is divided into 2 parts: a technical evaluation of RT-PCR assays used for detection of prostate cells and a discussion of the results of clinical studies using these procedures. REVERSE TRANSCRIPTASE-POLYMERASE CHAIN REACTION RT-PCR is a widespread technique for producing easily detect- able amounts of DNA from very small starting samples of RNA (Doherty et al., 1989). The method is highly suitable for detecting the presence of cells that express a message of interest, because the mRNA content of a single cell is often sufficient to yield a signal. The premise of the RT-PCR approach is that the cells to be detected express an mRNA that is not found in other cells within the environment being tested. The process of developing an RT-PCR protocol thus begins with the identification of an appropriate target gene. Even if the transcript is expressed by other cell types within the body, it may be a suitable target as long as it is not normally expressed within the environment being tested (e.g., cytokeratin [Schoenfeld et al., 1997]). Since 1987, RT-PCR methodology has been used in detecting circulating cells shed by solid tumors (see Pelkey et al., 1996, for review), utilizing, for example, tyrosinase for melanoma (Wang et al., 1994); tyrosine hydroxylase for neuroblastoma (Miyajima et al., 1995); alpha fetal protein for hepatoma (Komeda et al., 1995); and cytokeratin for breast cancer (Wernert et al., 1992). This broad spectrum shows the diagnostic potential of the RT-PCR method. Vessella et al. (1992) and Moreno et al. (1992) pioneered the use of RT-PCR for detection of migrating prostate cells. Since these initial reports, which made use of the PSA message, a number of other messages have been tested; the current target transcripts of choice for detection of prostate cells are PSA (Lundwall, 1989; Riegman et al., 1988), PSMA (Fair et al., 1997; Israeli et al., 1994b) and hK2 (Schedlich et al., 1987). Sources of RNA and RNA isolation Dissemination of cancerous cells to targeted organs probably occurs via PB and/or the lymphatic system. PB is an attractive source of samples for detection of disseminated cells, because of the ease of obtaining blood samples from patients and the high probability that the bloodstream is a critical route of metastatic migration. Although metastases are often found in LN of CaP patients, these samples are obtained only as a result of surgery; studies using LN have therefore been performed much less frequently. Finally, the BM, which is the major metastatic site of CaP, is of great interest as a source of disseminated prostate cells. BM aspirates can be obtained from patients in clinics or immedi- ately before radical prostatectomy (RP). Grant sponsor: Richard M. Lucas Foundation; Grant sponsor: George M. O’Brien Center Award from NIDDK; Grant number: 1-P50 DK/CA47656–03. *Correspondence to: Urology Department, Mail Stop 356510, the School of Medicine of the University of Washington, Seattle, WA 98195, USA. Fax: (206) 543–1146. E-mail: ecorey@u.washington.edu Received 3 February 1998; Revised 14 March 1998 Int. J. Cancer: 77, 655–673 (1998)  1998 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer