ELSEVIER DETECTION OF CIRCULATING PROSTATE CELLS BY REVERSE TRANSCRIPTASE-POLYMERASE CHAIN REACTION OF HUMAN GLANDULAR KALLIKREIN (hK2) AND PROSTATE-SPECIFIC ANTIGEN (PSA) MESSAGES EVA COREY, EDWARD W. ARFMAN, MATTHEW M. OSWIN, SEBASTIAN W. MELCHIOR, DONALD J. TINDALL, CHARLES Y.-F. YOUNG, WILLIAM J. ELLIS, AND ROBERT L. VESSELLA ABSTRACT Objectives. To investigate the clinical value of human glandular kallikrein (hK2) reverse transcriptase-poly- merase chain reaction (RT-PCR] for detection of prostate cells in circulation and to compare the results with those obtained from prostate-specific antigen (PSA) RT-PCR. Methods. We examined peripheral blood (PB) and bone marrow (BM) samples of 13 patients with advanced- stage prostate cancer and 63 patients with clinically localized disease for the presence of circulating prostate cells. An RT-PCR protocol with a two-step amplification cycle and hot-start conditions was used. Results. The limit of detection of the PCR portion is similar for PSA and hK2 (5 to 10 copies of the plasmid containing the cDNA). The RT-PCR limit of detection is one LNCaP cell in 1 O8peripheral blood mononuclear cells (PMBC) for PSA, and one LNCaP cell in 1 O7PMBC for hK2. Of the BM samples obtained prior to radical prostatectomy, 7 1.4% were positive for PSA mRNA and 41.3% were positive for hK2 mRNA. In PB, the PSA positivity was 19% and hK2 positivity 12.7%. In advanced-stage patients, there were 76.9% PSA-positive samples in BM versus 38.5% hK2-positive samples; 46.2% of patients were positive in PB for PSA versus 30.8% for hK2. Conclusions. We have developed a sensitive RT-PCR protocol for detection of hK2 mRNA and evaluated the suitability of hK2 mRNA in comparison with PSA mRNA as an additional marker for detection of prostate cells in circulation. Combining results of these two tests increased the sensitivity of detection. UROLOGY 50: 184-l 88, 1997. 0 1997, Elsevier Science Inc. All rights reserved. A bout 20% to 30% of patients diagnosed with clinically localized prostate cancer (Cap) suf- fer recurrence after radical prostatectomy. Disease recurrence may be caused by prostate cancer cells in bone marrow (BM) and peripheral blood (PB) that remained undetected at the time of diagnosis. Currently, reverse transcriptase-polymerase chain reaction (RT-PCR) procedures for prostate-spe- This work was funded by the Richard M. Lucas Foundation, the Department of Veterans Affairs, and a George M. O’Brien Center Award from NIDDK (1 P50 DK/CA47656-03). From the Tumor Immunology Laboratory of the Urology De- partment, School of Medicine of the University of Washington, Seattle, Washington; Department of Urology, Johannes Guten- berg University, Mainz, Germany; and Department of Urology, Mayo Clinic, Rochester, Minnesota Reprint requests: Eva Corey, Ph.D., Urology, Mail Stop 356510, University of Washington, Seattle, WA 98195 Submitted: March 25, 1997, accepted (with revisions): April 11,1997 0 1997, ELSEVIER SCIENCE INC. 184 ALLRIGHTS RESERVED cific antigen (PSA) and prostate-specific membrane antigen (PSMA) messenger ribonucleic acid (mRNA) are being evaluated for the detection of circulating prostate cells, which may prove to be correlated with metastatic spread of the disease. Despite enormous efforts devoted to this field, re- sults of these studies are still controversial. Because bone is the most common site of CaP metastasis, RT-PCR of BM specimens is of partic- ular interest for the detection of prostate cells. Wood et al.’ have sought a correlation between detection of PSA-positive cells by RT-PCR in BM and metastatic spread of Cap. We have shown that about 62% of BM specimens from patients under- going prostatectomy are positive for PSA mRNA.’ This percentage is higher than the percentage of patients who suffer recurrence. One possible ex- planation for the discrepancy is that not all cells that escape the prostate and are deposited in the BM are capable of causing metastasis. This would 0090-4295/97/$17.00 PII SOO90-4295(97)00262-S