Purpose: We determined whether circulating tumor cells predict prostate specific antigen failure in patients with metastatic prostate cancer before endocrine therapy and compared their prognostic ability with other clinical factors.
Materials and methods: Circulating tumor cells were enumerated with the CellSearchtrade mark system in whole blood. This system was developed using epithelial cell adhesion molecule antibody based immunomagnetic capture and automated staining methodology. Prostate cancer cell lines (PC3, LNCaP, DU145) and mixed blood from healthy men were analyzed using this system. Blood samples from 80 patients with metastatic prostate cancer before endocrine therapy were analyzed. Circulating tumor cells were then assessed every 3 months after endocrine therapy in these patients.
Results: Circulating tumor cell assay accuracy and reliability were determined using prostate cancer cell line (PC3, LNCaP, DU145) spiking experiments, which demonstrated a strong linear correlation (r = 0.99) and a constant recovery rate of 69% +/- 3%, 95% +/- 3% and 89% +/- 2%, respectively. The number of circulating tumor cells found ranged from 0 to 222 per 7.5 ml blood (mean 17 +/- 31, median 14). A threshold of 5 or more circulating tumor cells per 7.5 ml blood was used to evaluate the ability of circulating tumor cells to predict androgen deprivation responsiveness. Of the 80 patients 44 (55%) had 5 or more circulating tumor cells with a median androgen deprivation responsiveness of 17 months compared to more than 32 months for those with fewer than 5 circulating tumor cells (p = 0.007). The presence of circulating tumor cells, nadir prostate specific antigen values and Gleason score were significant parameters predictive of androgen deprivation responsiveness on univariate and multivariate analyses.
Conclusions: In this study the presence of 5 or more circulating tumor cells in 7.5 ml blood was associated with androgen deprivation responsiveness in patients with metastatic prostate cancer before endocrine therapy.