Prostate cancer is a significant health problem worldwide. Glucocorticoids, including' dexamethasone have formed the backbone of treatment of androgenindependent disease for the past 50 years and are given with docetaxel in the only therapy with a proven survival benefit in this setting. This study explored the mecha'nism of dexamethasone action in the treatment of androgen-independent prostate cancer (AIPC). In the in vitro study' dexamethasone administration induced a suppression of NFKB transcriptional activity in AIPC cells, resulting in a decreased transcription . arid/or secretion of the pro-angiogenic factors interleukin-8 and VEGF. In addition, co-administration of dexamethasone attenuated the docetaxel-induced .- '..: .. AP-1 and NF-KB transcriptional activity and abrogated the IL-8 gene transcription . - - -- - ·L and secretion in these cells and endothelial cells. In a novel finding, addition of dexamethasone potentiated the pronounced anti-angiogenic activity of docetaxel as assessed by an in vitro angiogenesis assay. Since no modulation of docetaxel cytotoxicity in endothelial cells by addition of. dexamethasone was observed, these results suggest that dexamethasone contributes to the efficacy of docetaxel through reduction of the docetaxel-induced increase of pro-angiogenic factors both in prostate cancer and endothelial cells, which translates into inhibition of angiogenesis. In the clinical study, the PSA response rate was 63%. Serum IL-8 was elevated in all 11 oatients for whom data is available. However. at this interim analvsis Supplied by The British Library - 'The world's knowledge' I I, I ' serum IL-8 levels were not consistently modified following dexamethasone therapy. However, when IL-8 is analysed as a binary variable, low IL-8 is associated with a prolonged time to biochemical progression which does not reach statistical significance at this time. Similarly, CRP levels were not affected consistently by dexamethasone but baseline CRP was lower in patients who achieved a PSA response than in those who did not respond, though this difference was not statistically significant.