Androgen Receptor Blockade in Experimental Combination Therapy of Pancreatic Cancer

Background: Reports on hormone receptor expression of pancreatic cancer (PaCa) cells and treatment responses to antihormonal therapy are still conflicting. Methods: Eight human PaCa cell lines were tested for androgen receptor (AR) protein levels by Western blot analysis. Cell proliferation in vitro was measured by sulforhodamine B analysis. AR agonists and inhibitors included dihydrotestosterone (DHT), testosterone (T), and flutamide (Flu). In vivo therapy of nude mouse xenografts tested Flu with gemcitabine (Gem) and/or bevacizumab (Bev). Results: Seven of eight human PaCa cell lines expressed detectable AR protein. Median relative expression compared with the AR positive control LnCaP was 21% (range: 16 to 63). Growth stimulation by DHT or T was minor (<20%); inhibition by Flu varied greatly and did not correlate to AR levels. Even in the sensitive cell line Panc1, Flu failed to increase Gem toxicity in vitro. However, in vivo Flu therapy resulted in significant growth inhibition of Panc-1 tumors. Flu/Gem treatment did not enhance the effect; Bev/Flu/Gem triple therapy had the greatest effect (P = 0.06 compared to Flu/Gem). Flu alone did not affect apoptotic activity, but decreased the tumor cell proliferative index (P = 0.04); in combination with Gem, Flu reduced the tumor cell density (P = 0.02). Conclusions: The majority of PaCa cell lines express AR at various levels, but most fail to show an in vitro antiproliferative response to AR inhibition. The strong antitumor effect of flutamide in vivo is not significantly enhanced in combination with gemcitabine or bevacizumab, suggesting primarily monotherapy benefit potential of AR blockade in susceptible PaCa.