Int. J. Cancer: 70, 52–56 (1997) r 1997 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer INCREASED ANALYTICAL SENSITIVITY OF RT-PCR OF PSA mRNA DECREASES DIAGNOSTIC SPECIFICITY OF DETECTION OF PROSTATIC CELLS IN BLOOD Wolfgang HENKE1*, Monika JUNG1, Klaus JUNG1, Michael LEIN1, Horst SCHLECHTE1, Christoph BERNDT2, Birgit RUDOLPH3, Dietmar SCHNORR1 and Stefan A. LOENING 1Klinik und Poliklinik für Urologie, Universitätsklinikum Charité, Medizinische Fakultät der Humboldt Universität zu Berlin, D-10098 Berlin, Germany 2Institut für Pathologische und Klinische Biochemie, Universitätsklinikum Charité, Medizinische Fakultät der Humboldt Universität zu Berlin, D-10098 Berlin, Germany 3Institut für Pathologie, Universitätsklinikum Charité, Medizinische Fakultät der Humboldt Universität zu Berlin, D-10098 Berlin, Germany The diagnostic specificity of the detection of disseminated prostatic cells by reverse-transcriptase polymerase chain reaction (RT-PCR) of PSA mRNA was investigated. A sensitive nested PCR was developed. In blood samples from 10 healthy female and 10 healthy male persons examined by RT-PCR, mRNA of PSA was detected 3 times in each group. In the groups of patients suffering from benign prostate hyperplasia and prostate cancer, 6 of 11 and 5 of 12, respectively, gave positive RT-PCR results. With increasing analytical sensitivity of the RT-PCR of PSA mRNA, the diagnostic specificity of the assay is decreased. Further development of this diagnostic method requires the introduction of the quantitative PCR which may make possible discrimination between prostatic and non-prostatic source of PSA mRNA by quantification. Int. J. Cancer, 70:52–56, 1997. r 1997 Wiley-Liss, Inc. Early detection of metastases is of great relevance for the therapy of prostate cancer. Recently, the detection of mRNA of prostatespecific antigen (PSA) by reverse-transcriptase polymerase chain reaction (RT-PCR) has been introduced to identify prostatic cells in blood (Moreno et al., 1992), lymph nodes (Deguchi et al., 1993) and bone marrow (Wood et al., 1994). Provided that the amplified mRNA is expressed exclusively in the cell of interest, its detection outside the original organ identifies disseminated cells. PSA is thought to be expressed only in the prostatic cell (Oesterling, 1991) and thus fulfills the requirement of this method. The expression of PSA is not tumor-specific, so the malignancy or metastatic power of the disseminated cell cannot be evidenced. Authors using PSA mRNA as an indicator of prostatic cells disseminated in blood obtained positive RT-PCR results in the range from 25% (Israeli et al., 1994a) to 80% (Cama et al., 1995) of patients with metastatic disease. The diversity of published results indicates that either the blood of some patients with metastatic disease does not contain PSA-expressing cells or the sensitivity of the applied method is limited. Nevertheless, RT-PCR allows the study of such interesting questions as the intra-operative spillage and hematogenous dissemination of prostate-cancer cells during radical prostatectomy (Eschwège et al., 1995; Oefelein et al., 1996) or the prediction of treatment failure following radical prostatectomy (Olsson et al., 1996). The diagnostic power of RT-PCR of PSA mRNA depends on the reliability of the applied method. Therefore, we are interested in increasing the sensitivity of this approach. The present study demonstrates that increased analytical sensitivity of RT-PCR leads to the detection of PSA mRNA in the blood of healthy females and males. MATERIAL AND METHODS Subjects We studied 43 subjects, 10 of whom were healthy females and 10 healthy males; 11 were patients with benign prostatic hyperplasia and 12 were patients with prostate cancer without additional disease. The average ages (median) of the blood donors, the patients with hyperplasia and the cancer patients were 43, 67 and 64 respectively. The diagnosis of benign prostate hyperplasia was based in 6 cases on histological analysis of tissue obtained by transurethral resection of the gland and in 5 cases on conventional clinical data (digital rectal examination, transrectal sonogram). The cancer diagnosis was verified by histology of surgically removed specimens (6 cases were stages T1 to T3N0M0, and 6 were T2N1M0 to T4N2M0). Isolation of mononucleated cells from blood and RNA extraction Blood samples were collected in 10 ml EDTA-K monovettes (Sarstedt, Nümbrecht, Germany) at 1 week pre-operatively and at least 3 weeks after digital rectal examination or prostate needle biopsy. Mononucleated cells were harvested from blood by FicollPaque (Pharmacia, Uppsala, Sweden) gradient centrifugation. Total RNA was isolated by the guanidinium-thiocyanate/phenol/ chloroform extraction method (Chomczynski and Sacchi, 1987) using the TRIzol reagent (GIBCO BRL, Eggenstein, Germany). The RNA pellet obtained after precipitation with isopropanol and washing with ethanol was allowed to air-dry and then re-dissolved in 0.1% diethylpyrocarbonate-treated water. The RNA content was measured spectrophotometrically at 260 nm. Reverse transcription Total RNA (1 µg) was used for cDNA synthesis, applying the SuperScript II reverse-transcriptase assay (GIBCO BRL) according to the manufacturer’s instructions. RNA and 0.5 µg oligo(dT)12–18 primers were first denaturated for 10 min at 70°C, chilled on ice for 1 min, and then incubated for 50 min at 42°C in 20 µl of a reaction mixture containing 20 mM Tris-HCl buffer, pH 8.4, 50 mM KCl, 2.5 mM MgCl2, 10 mM dithiothreitol, 0.5 mM deoxynucleoside triphosphate mix (dNTP), and 200 U of SuperScript II reverse transcriptase. The reaction was terminated by heating for 15 min at 70°C. The assay was completed by incubation with RNase H for 20 min at 37°C. PCR The synthesized cDNA (10 µl) were used in PCR. The 50-µl PCR assay contained the following components: 16.7 mM (NH4)2SO4, 67 mM Tris-HCl buffer, pH 8.8, 0.02% (w/v) gelatine, 10 mM mercaptoethanol, 100 µM dNTP mix and 2.5 U Taq DNA polymerase (Promega, Heidelberg, Germany). The mixture was overlaid with mineral oil (Sigma, Deisenhofen, Germany), heated in the UNO thermal cycler (Biometra, Göttingen, Germany) at 94°C for 4 min and exposed to 40 cycles of heating at 48°C for 1 min, at 72°C for 1 min and at 94°C for 1 min. The last cycle was *Correspondence to: Universitätsklinikum Charité, Medizinische Fakultät der Humboldt-Universität zu Berlin, Forschungsabteilung der Klinik für Urologie, Schumannstrasse 20–21, D-10098 Berlin, Germany. Fax: (049) 30/28021402. E-mail: Henke@rz.charite.hu-berlin.de. Received 6 June 1996; revised 5 September 1996. DETECTION OF PROSTATIC CELLS IN BLOOD BY RT-PCR 53 FIGURE 1 – Scheme of primer localization on PSA mRNA. Symbols: M, I, K and W, primers published by Moreno et al. 1992), Israeli et al. 1994a), Katz et al. (1994) and Wood et al. 1994); subscripts s and a, sense and anti-sense primer. finished after prolonged annealing at 72°C for 3 min by cooling to 4°C. Then 3 µl of the first PCR were employed for the second PCR, which followed the protocol of the first round. Primers The primer sets published by Moreno et al. (1992) and Wood et al. (1994) as well as the outer primer sets of Israeli et al. (1994a) and Katz et al. (1994) were included in our investigations. Figure 1 illustrates the binding sites of these primers on exons 3, 4, and 5 of PSA mRNA. Finally, we tested different primer combinations with nested PCR. PSA displays a high degree of homology with human glandular kallikrein. The published primers are selected to avoid co-amplification of this prostatic mRNA. b-globin primers were used to verify the integrity of RNA as described (Jaakkola et al., 1995). Agarose-gel electrophoresis, digestion by SacI, and sequencing The PCR products were separated by agarose-gel electrophoresis using 2% agarose (Nusieve, FMC Bioproducts, Rockland, ME) and visualized with ethidium bromide. The PCR product was digested with the restriction enzyme SacI (Promega) according to the manufacturer’s instructions. PCR products were sequenced by the dideoxynucleotide-chaintermination method on the automated DNA sequencer A.L.F. (Pharmacia, Freiburg, Germany). The second PCR round was performed with a biotinylated primer, used to separate the NaOHtreated double DNA strand by streptavidin-coated magnetic beads (Dynal, Hamburg, Germany), and a fluorescein-labeled sequencing primer nested in a 3-base-pair distance to the original primer. The sequencing reaction was done by the sequenase system. Cell culture LNCaP cells were grown in RPMI-1640 medium supplemented with 10% FCS, 125 kU penicillin, and 125 mg/l streptomycin. Confluent monolayers were washed twice with 2 ml of PBS at room temperature and detached using trypsin/EDTA solution (0.05%, 0.02%) at 37°C. Serial dilutions of LNCaP cells in PBS were added to 5 ml blood of female subjects, mixed gently and treated as described above. RESULTS We investigated detection of the mRNA of PSA with a nested PCR consisting of 40 cycles per each round. To optimize the sensitivity of this assay, several combinations of published primers were tested. Finally, the primer combination 1, 2, and 3 consisting of Is/Ia and Ms/Ma, Is/Ka and Ks/Ia, and Ks/Ka and Ws/Wa were investigated in the first and second rounds of PCR (Fig. 1). The assay sensitivity was determinated by dilution series of PSAgenerating LNCaP cells using blood of healthy females. Whereas FIGURE 2 – Sensitivity of the detection of prostatic cells by RT-PCR of PSA mRNA tested by dilution series of LNCAP cells in whole blood. Lanes 3–5, 6–8 and 9–11, primer combinations 1, 2 and 3, respectively, using 50, 5 and 1 LNCaP cells/106 leukocytes (ascending lane number); lane 1, water control; lanes 2 and 12, DNA marker IX (Boehringer, Mannheim, Germany). TABLE I – DETECTION OF PSA mRNA BY NESTED RT-PCR IN BLOOD OF HEALTHY FEMALE AND MALE SUBJECTS AND IN PATIENTS SUFFERING FROM BENIGN PROSTATE HYPERPLASIA AND PROSTATE CANCER Source Number Positive RT-PCR results (%) Healthy females Healthy males Benign prostate hyperplasia patients Prostate-cancer patients 10 10 11 12 3 (33) 3 (33) 6 (54) 5 (42) an average of 1 LNCaP cell per 106 leukocytes is detected by using primer combination 1, primer combinations 2 and 3 give lower sensitivity, about 5 and 50 LNCaP cells per 106 leukocytes respectively (Fig. 2). Further investigations were performed with primer combination 1. Analyzing blood from 10 healthy female and 10 healthy male subjects by nested RT-PCR, we detected PSA mRNA 3 times in each group (Table I). In the groups of patients suffering from benign prostate hyperplasia and prostate cancer, 6 of 11 and 5 of 12 gave positive RT-PCR results (Fig. 3). Controls performed without added RNA or reverse transcriptase were always negative, so it is unlikely that the positive results are caused by cross-contamination. Since the primers of the second PCR bind in exon 3 and exon 4, the amplicon of PSA cDNA can be clearly distinguished from that of the intron-containing genomic DNA (Moreno et al., 1992), which might contaminate the RNA. The amplicon of PSA has a cleavage site for the restriction enzyme SacI. Digestion of the amplicon with SacI produced 2 restriction fragments of the expected size (Fig. 4). Human glandular kallikrein cDNA, the sequence of which has a high degree of homology with that of PSA, has no SacI restriction site. Thus, the amplification of human glandular kallikrein mRNA is ruled out. The identity of the desired PCR products was verified by sequencing the amplicon by the dideoxynucleotide-chain-termination method with a fluorescein-labeled primer using the sequenase reaction. The determined sequences of the amplicons from one healthy female and one healthy male, as well as from one HENKE ET AL. 54 TABLE II – SUMMARY OF PUBLISHED RESULTS OBTAINED WITH RT-PCR OF PSA AND PSM IN BLOOD OF PATIENTS WITH METASTATIC PROSTATE CANCER Author mRNA Moreno et al. (1992) Seiden et al. (1994) Katz et al. (1994) Jaakkola et al. (1995) Ghossein et al. (1995) Cama et al. (1995) PSA PSA PSA PSA PSA PSA PSM PSA PSM PSA PSM Israeli et al. (1994a) Loric et al. (1995) FIGURE 3 – Detection of PSA mRNA in blood of persons without prostate cancer. Lanes 2 to 6, PSA primers; lanes 6 and 11, water control; lanes 7 to 11, b-globin primers; lanes 1 and 12, DNA marker IX; lane 2, male subject; lanes 3 and 4, female subject; lane 5, patient with benign prostate hyperplasia. FIGURE 4 – Digestion of PCR product by the restriction enzyme SacI. PCR product of a healthy female before and after digestion with SacI (lanes 1 and 2). DNA marker IX (lane 3) and water control (lane 4). benign-prostate-hyperplasia patient, agree with those published for PSA by Lundwall (1989). DISCUSSION Our data show that mRNA of PSA can be detected in the blood of healthy female and male subjects with a sensitive RT-PCR applying nested primers. Data from controls make it unlikely that the positive results are caused by cross-contamination, human glandularkallikrein cDNA or genomic DNA. Our results are in contrast to data of several other studies (Moreno et al., 1992; Katz et al., 1994; Seiden et al., 1994; Ghossein et al., 1995; Israeli et al., 1995; Jaakkola et al., 1995; Oefelein et al., 1996). However, Smith et al. (1995) and O’Hara et al. (1996) also detected PSA mRNA by a Patients with metastatic prostate cancer 12 35 18 18 76 20 24 33 Positive RT-PCR results (%) 4 (33) 11 (31) 14 (78) 9 (50) 26 (34) 16 (80) 10 (50) 6 (25) 16 (67) 17 (51) 28 (85) sensitive nested PCR in non-prostatic cell lines and in blood from healthy men and women. Zarghami and Diamandis (1996) found PSA mRNA in breast tumors of female patients. From the published data, we conclude that the analytical sensitivity of the RT-PCR is one determinant for the diagnostic significance of this assay. The sensitivity of our assay was optimized by an increased number of cycles per PCR and selection of one from 3 primer combinations investigated. Primer combination 1, incorporating primer sets published by Israeli et al. (1994a) and Moreno et al. (1992), gave the highest sensitivity in our experimental conditions. The finding that primer selection is decisive for the sensitivity of PCR is in agreement with the experience of He et al. (1994). We evaluated the assay sensitivity by dilution series of LNCaP cells with blood and detected an average of 1 LNCaP cell per 106 leukocytes. With whole blood used as diluting medium, the same sensitivity has been also found by other authors (Seiden et al., 1994; Jaakkola et al., 1995; Loric et al., 1995). In several reports, the assay sensitivity was measured by diluting LNCaP cells with isolated non-prostatic cells (Deguchi et al., 1993; Israeli et al., 1994a,b; Katz et al., 1994; Wood et al., 1994; Ghossein et al., 1995). The sensitivity determined in this way can hardly be compared with that based on whole blood, since recovery of LNCaP cells in the mononuclear fraction was not evaluated. The dilution of cDNA (Israeli et al., 1994a) with water or of RNA isolated from LNCaP cells with non-prostatic RNA (Seiden et al., 1994; Oefelein et al., 1996) has no practical relevance for the characterization of the assay sensitivity. Table II summarizes results obtained with RT-PCR of PSA and prostate-specific membrane antigen (PSM), a further protein assumed to be specifically expressed in prostatic cells (Israeli et al., 1994b), from blood of patients with prostate cancer. Israeli et al. (1994a) or Loric et al. (1995), targeting on PSA, detected prostate cells in only 25 or 51%, respectively, of patients with metastatic cancer. However, prostate cells are present in blood, as can be concluded from the data obtained with RT-PCR of mRNA of PSM (Israeli et al., 1994a; Loric et al., 1995), of at least 67 or 85% of the same patients. Obviously, the RT-PCR of PSA mRNA performed by these authors does not detect all cases of disseminating prostate cells. The RT-PCR applied by Cama et al. (1995) seems to be characterized by an inverse relationship of the sensitivity for mRNA of PSA and PSM, since their methods detected prostate cells in 80 and 50% of the investigated patients respectively. The authors, investigating mRNA of both PSA and PSM, checked the sensitivity of the respective method by dilution series of LNCaP cells. No sensitivity differences are described for the 2 targets (Israeli et al., 1994a; Cama et al., 1995; Loric et al., 1995). Thus, the method currently used to evaluate the sensitivity gives only a rough estimation of the analytical sensitivity. Eschwège et al. (1995) investigated the hematogenous dissemination of prostatic cells during radical prostatectomy by RT-PCR of DETECTION OF PROSTATIC CELLS IN BLOOD BY RT-PCR PSM mRNA and found that anti-androgen therapy reduces the hematogenous spread. Hamdy and Neal (1996) pointed out that androgen ablation depressed the number of tumor cells by apoptosis. Therefore, the drop in numbers of disseminating prostate cells below the assay sensitivity should be considered as one possible reason for the finding of Eschwège et al. (1995). Olsson et al. (1996) evaluated the role of RT-PCR of PSA mRNA in predicting disease recurrence. Comparing the predictive power of pre-operative serum PSA, Gleason score and PSA mRNA, the RT-PCR assay has the lowest univariate relative risk. The applied RT-PCR method gave positive results in 85% of patients with confirmed metastatic prostate cancer. In the light of published data, the sensitivity of the RT-PCR should be increased, to reduce the influence of methodical uncertainty on the result of studies. Diamandis and Yu (1995) pointed to several methodological facts that could explain the requirement to increase the analytical sensitivity of this approach. Such an increase of analytical sensitivity by applying a nested PCR implies, in agreement with Smith et al. (1995) and O’Hara et al. (1996), with a decrease in diagnostic specificity. The assumption that PSA is specifically expressed in prostate cells is now being revised. Thus, immunoreactive PSA is ultrasensitively-assayed in normal endometrial tissue and amniotic fluid, also in breast, ovarian, liver, kidney, adrenal, colon, parotid and lung tumors (Diamandis, 1995). More than one thousand female sera were analyzed with an ultrasensitive PSA assay. In 17% of persons in this group, PSA was measurable (Yu and Diamandis, 1995). Randell et al. (1996) found detectable concentrations of PSA in many cord sera and in sera 55 from male and female neonates. Although most children at ages 1 to 10 years had undectable PSA, it was found in a few. The lack of any apparent sex-related difference again suggests that nonprostatic sources may contribute to serum PSA (Randell et al., 1996). Knowledge of the biological functions of PSA, originally attributed solely to the liquefaction of seminal fluid, is increasing. Published results suggest that PSA is either a growth factor or a growth-factor regulator (Diamandis, 1995). Besides growing evidence of PSA expression in non-prostatic tissues, some data show that increased analytical sensitivity of PCR may be connected with decreased diagnostic specificity. Schoenfeld et al. (1994) tested RT-PCR of cytokeratin 19 mRNA for detecting breast-cancer micrometastases in axillary lymph nodes. With a sensitivity increased by nested PCR, this mRNA was found in lymph nodes without cancer. Biernaux et al. (1995) found mRNA of the major bcr-abl fusion gene, presently seen as the hallmark of chronic myeloid leukemia, in the blood of healthy individuals, by means of a very sensitive nested RT-PCR. Using RT-PCR of PSM mRNA positive results with healthy female and male subjects are not described in the literature. Further development of this assay might therefore be concentrated on PSM mRNA as an indicator of disseminating prostate cells. Nevertheless, the introduction of competitive RT-PCR should improve the reliability of this method, since the assay sensitivity can be controlled and the signal intensity can be quantified, e.g., by a microtiter-plate-based assay (Galvan et al., 1995). Quantification might make possible discrimination between non-prostatic and prostate-derived mRNA. REFERENCES BIERNAUX, C., LOOS, M., SELS, A., HUEZ, G. and STRYCKMANS, P., Detection of major bcr-abl gene expression at a very low level in blood cells of some healthy individuals. Blood, 86, 3118–3122 (1995). CAMA, C., OLSSON, C.A., RAFFO, A.J., PERLMAN, H., BUTTYAN, R., O’TOOLE, K., MCMAHON, D., BENSON, M.C. and KATZ, A.E., Molecular staging of prostate cancer. II. A comparison of the application of an enhanced reverse-transcriptase-polymerase-chain-reaction assay for prostatespecific antigen versus prostate-specific membrane antigen. J. Urol., 153, 1373–1378 (1995). CHOMCZYNSKI, P. and SACCHI, N., Single-step method of RNA isolation by acid-guanidinium-thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162, 156–159 (1987). DEGUCHI, T. and 13 OTHERS, Detection of micrometastatic prostate cancer cells in lymph nodes by reverse-transcriptase-polymerase-chain reaction. Cancer Res., 53, 5350–5354 (1993). DIAMANDIS, E.P., New diagnostic applications and physiological functions of prostate-specific antigen. Scand. J. clin. Lab. Invest., Suppl., 221, 105–112 (1995). DIAMANDIS, E.P. and YU, H., Prostate cancer, prostate-specific antigen, and the polymerase chain reaction. Clin. Chem., 41, 177–179 (1995). ESCHWÈGE, P., DUMAS, F., BLANCHET, P., LE MAIRE, V., BENOIT, G., JARDIN, A., LACOUR, B. and LORIC, S., Haematogenous dissemination of prostatic epithelial cells during radical prostatectomy (comments). Lancet, 346, 1528–1530 (1995). GALVAN, B., CHRISTOPOULOS, T.K. and DIAMANDIS, E.P., Detection of prostate-specific antigen mRNA by reverse-transcription-polymerase-chain reaction and time-resolved fluorometry. Clin. Chem., 41, 1705–1709 (1995). GHOSSEIN, R.A., SCHER, H.I., GERALD, W.L., KELLY, W.K., CURLEY, T., AMSTERDAM, A., ZHANG, Z.F. and ROSAI, J., Detection of circulating tumor cells in patients with localized and metastatic prostatic carcinoma: clinical implications. J. clin. Oncol., 13, 1195–1200 (1995). HAMDY, F.C. and NEAL, D.E., Haematogenous dissemination of prostate epithelial cells during surgery (letter). Lancet, 347, 325 (1996). HE, Q., MARJAMAKI, M., SOINI, H., MERTSOLA, J. and VILJANEN, M.K., Primers are decisive for sensitivity of PCR. Biotechniques, 17, 82, 84, 86–7 (1994). ISRAELI, R.S., MILLER, W.H., JR., SU, S.L., POWELL, C.T., FAIR, W.R., SAMADI, D.S., HURYK, R.F., DEBLASIO, A., EDWARDS, E.T., WISE, G.J. and HESTON, W.D., Sensitive nested reverse-transcription-polymerase-chain- reaction detection of circulating prostatic tumor cells: comparison of prostate-specific membrane antigen and prostate-specific antigen-based assays. Cancer Res., 54, 6306–6310 (1994a). ISRAELI, R.S., MILLER, W.H., JR., SU, S.L., SAMADI, D.S., POWELL, C.T., HESTON, W.D., WISE, G.J. and FAIR, W.R., Sensitive detection of prostatichematogenous-tumor-cell dissemination using prostate-specific antigen and prostate-specific membrane-derived primers in the polymerase chain reaction. J. Urol., 153, 573–577 (1995). ISRAELI, R.S., POWELL, C.T., CORR, J.G., FAIR, W.R. and HESTON, W.D., Expression of the prostate-specific membrane antigen. Cancer Res., 54, 1807–1811 (1994b). JAAKKOLA, S., VORNANEN, T., LEINONEN, J., RANNIKKO, S. and STENMAN, U.H., Detection of prostatic cells in peripheral blood: correlation with serum concentrations of prostate-specific antigen (comments). Clin. Chem., 41, 182–186 (1995). KATZ, A.E., WIELOPOLSKI, L., RAFFO, A.J., CAMA, C., PERLMAN, H., SEAMAN, E., O’TOOLE, K.M., MCMAHON, D., BENSON, M.C. and BUTTYAN, R., Molecular staging of prostate cancer with the use of an enhanced reverse-transcriptase-PCR assay. Urology, 43, 765–775 (1994). LORIC, S., DUMAS, F., ESCHWÈGE, P., BLANCHET, P., BENOIT, G., JARDIN, A. and LACOUR, B., Enhanced detection of hematogenous circulating prostatic cells in patients with prostate adenocarcinoma by using nested reversetranscription-polymerase-chain-reaction assay based on prostate-specific membrane antigen (comments). Clin. Chem., 41, 1698–1704 (1995). LUNDWALL, A., Characterization of the gene for prostate-specific antigen, a human glandular kallikrein. Biochem. biophys. Res. Commun., 161, 1151– 1159 (1989). MORENO, J.G., CROCE, C.M., FISCHER, R., MONNE, M., VIHKO, P., MULHOLLAND, S.G. and GOMELLA, L.G., Detection of hematogenous micrometastasis in patients with prostate cancer. Cancer Res., 52, 6110–6112 (1992). OEFELEIN, M.G., KAUL, K., HERZ, B., BLUM, M.D., HOLLAND, J.M., KEELER, T.C., COOK, W.A. and IGNATOFF, J.M., Molecular detection of prostate epithelial cells from the surgical field and peripheral circulation during radical prostatectomy (comments). J. Urol., 155, 238–242 (1996). OESTERLING, J.E., Prostate-specific antigen: a critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J. Urol., 145, 907–923 (1991). O’HARA, S.M., VELTRI, R.W., SKRIPSTUNAS, P., HEDICAN, S.P., PARTIN, A.W., NELSON, J.B., SUBONG, E.N.P. and WALSH, P.C., Basal PSA mRNA levels detected by quantitative reverse-transcriptase-polymerase- 56 HENKE ET AL. chain reaction (Q-RT-PCR-PSA) in blood from subjects without prostate cancer. J. Urol., 155, (Suppl.) 418A (1996). OLSSON, C.A., DE VRIES, G.M., RAFFO, A.J., BENSON, M.C., O’TOOLE, K., CAO, Y., BUTTYAN, R.E. and KATZ, A.E., Pre-operative reverse-transcriptasepolymerase-chain reaction for prostate-specific antigen predicts treatment failure following radical prostatectomy. J. Urol., 155, 1557–1562 (1996). RANDELL, E.W., DIAMANDIS, E.P. and ELLIS, G., Serum prostate-specific antigen measured in children from birth to age 18 years. Clin. Chem., 42, 420–423 (1996). SCHOENFELD., A., LUQMANI, Y., SMITH, D., O’REILLY, S., SHOUSHA, S., SINNETT, H.D. and COOMBES, R.C., Detection of breast-cancer micrometastases in axillary lymph nodes by using polymerase chain reaction. Cancer Res., 54, 2986–2990 (1994). SEIDEN, M.V., KANTOFF, P.W., KRITHIVAS, K., PROPERT, K., BRYANT, M., HALTOM, E., GAYNES, L., KAPLAN, I., BUBLEY, G., DEWOLF, W. and SKLAR, J., Detection of circulating tumor cells in men with localized prostate cancer. J. clin. Oncol., 12, 2634–2639 (1994). SMITH, M.R., BIGGAR, S. and HUSSAIN, M., Prostate-specific antigen messenger RNA is expressed in non-prostate cells: implications for detection of micrometastases. Cancer Res., 55, 2640–2644 (1995). WOOD, D.P.J., BANKS, E.R., HUMPHREYS, S. and RANGNEKAR, V.M., Sensitivity of immunohistochemistry and polymerase chain reaction in detecting prostate-cancer cells in bone marrow. J. Histochem. Cytochem., 42, 505–511 (1994). YU, H. and DIAMANDIS, E.P., Measurement of serum prostate-specific antigen levels in women and in prostatectomized men with an ultrasensitive immunoassay technique. J. Urol., 153, 1004–1008 (1995). ZARGHAMI, N. and DIAMANDIS, E.P., Detection of prostate-specific antigen mRNA and protein in breast tumors. Clin. Chem., 42, 361–366 (1996).