Presentation Authors: Christian Pavlovich*, Jillian Eskra, Daniel Rabizadeh, Jiayi Zhang, William Isaacs, Jun Luo, Baltimore, MD
Introduction: Cells of prostatic origin can be detected in urine samples from men with prostate cancer, though typically at low frequency. Cellular exfoliation is enhanced by digital rectal examination (DRE), and by the utilization of molecular detection methods. To overcome technical challenges in urinary prostate cancer cell detection, we have developed a novel RNA in situ hybridization assay (RISH) for use on post-DRE urine sediments.
Methods: First-voided urine was collected after DRE in 98 patients presenting for prostate biopsy. Urine cytology slides were prepared for detection of rare prostate cancer cells by a validated multiplex RISH probe-set consisting of NKX3.1, PRAC1, and PCA3. Each patient was assigned to one of four biomarker status groups: 1) NKX3.1/PRCA1+ and PCA3+ (prostate cancer cells); 2) NKX3.1/PRCA1+ and PCA3- (prostate cells); 3) NKX3.1/PRCA1- and PCA3- (no prostate cells); and 4) uncertain status. Biomarker calls were made blind to clinical data. We evaluated the correlation between biomarker status and cancer status in this pilot all-comer cohort using a pre-defined, yet preliminary analysis plan.
Results: Of the 98 patients tested by the urine RISH assay, 30 had no prostate cancer, and 68 were diagnosed with prostate cancer, including 48 with Epstein criteria clinically significant disease (csPCa). Cells of prostatic origin (NKX3.1/PRCA1+ and PCA3+ or PCA3-) were detected in 69 patients, including 20 that had prostate cancer cells identified in their post-DRE urine sediment (NKX3.1/PRCA1+ and PCA3+). Nine patients were negative for urinary cells of prostatic origin, and the remaining 20 patients had indeterminate biomarkers status, mainly due to cell crowding rendering indeterminate biomarker calls. Among the 20 patients in the (NKX3.1/PRCA1+ and PCA3+) group, 18 had a positive biopsy, and all 18 were positive for csPCa; the other 2 were negative on biopsy. Within the biopsy-positive cohort overall (n=68), a positive detection correlated with high-risk cancer features (p=0.036), PSA density (p=0.022), and csPCa (p < 0.0001), but not with Gleason score alone (p=0.172) or total PSA alone (p=0.365). Excluding the indeterminate patients, the novel RISH test was 95% specific and 51% sensitive for the detection of csPCa.
Conclusions: Using a novel multiplex urine RISH test, cells of prostatic origin that appear benign or malignant can be definitively detected and visualized. The detection of prostate cancer cells by this approach is highly specific for the detection of csPCa. Future studies in a larger cohort will help optimize and validate the clinical utility of this assay.
Source of Funding: Patrick C. Walsh Foundation