Evaluation of vitamin D analogs as therapeutic agents for prostate cancer.

Prostate cancer cells contain specific receptors (VDR) for la,25-dihydroxyvitamin D (1alpha,25(OH)2D), which is known to inhibit the proliferation and invasiveness of these cells. These findings support the use of 1alph,25(OH)2D for prostate cancer therapy. However, because 1alpha,25(OH)2D can cause hypercalcemia, analogs of 1alpha,25(OH)2D that are less calcemic but which exhibit potent antiproliferative activity would be attractive as therapeutic agents. We studied four vitamin D compounds: 25-hydroxyvitaminD3 [25(OH)D3], which is converted to 1alpha,25(OH)2D3 in prostate cells, and three analogs of 1alpha,25(OH)2D3: EB1089, 19-nor-1alpha,25(OH)2D2 and hexafluoro-1alpha,25(OH)2D3 (F6-1alpha,25(OH)2D3). 19-nor-1alpha,25(OH)2D2 has been shown to be less calcemic than 1alpha,25(OH)2D3 in clinical trials. F6-1alpha,25(OH)2D3 has been shown to be 100-fold more active than 1alpha,25(OH)2D3 and to be longer-lasting in inhibiting keratinocyte proliferation in vitro. EB1089 has been shown to be less calcemic than 1alpha,25(OH)2D3 in rats implanted with Leydig cell tumors. For 25(OH)D3, 19-nor-1alpha,25(OH)2D2 and F6-1alpha,25(OH)2D3, we studied the in vitro effects and compared their activity to 1alpha,25(OH)2D3 on cellular proliferation by 3H-thymidine incorporation assay. In addition, we studied transactivation of the VDR in the presence of 25(OH)D3 and 19-nor-1alpha,25(OH)2D2 in prostate cells. For EB1089, we compared its inhibition of prostate cancer metastasis to that induced by 1alpha,25(OH)2D3 in vivo in the rat Dunning MAT LyLu prostate cancer model. We found that 1alpha,25(OH)2D3 and 19-nor-1alpha,25(OH)2D2 caused similar dose-dependent inhibition in 3H-thymidine incorporation into DNA in prostate cells and behaved similarly in the CAT reporter gene transactivation assay in PC-3/VDR cells. F6-1alpha,25(OH)2D3 is 10- to 50-fold more active than 1alpha,25(OH)2D3 in 3H-thymidine incorporation into DNA in the primary cultured prostate cells. Likewise, 25(OH)D3 had comparable antiproliferative activity to la,25(OH)2D3. In the rat model, tumor volumes and the number of metastases in the lungs were significantly reduced by both 1alpha,25(OH)2D3 (10.4 +/- 2.81 tumor foci) and EB1089 (7.7+/-1.29 tumor foci) compared to controls (22.7 +/- 1.98 tumor foci). Although serum calcium levels were significantly elevated in both 1alph,25(OH)2D3- and EB1089-treated rats, EB1089 was significantly less calcemic than 1alpha,25(OH)2D3 (12.59+/-0.21 mg/dl versus 14.47+/-.46 mg/dL; 1 microg/kg; p < 0.001). In conclusion, our data indicate that 25(OH)D3 and the three 1alpha,25(OH)2D analogs represent two different solutions to the problem of hypercalcemia associated with vitamin D-based prostate cancer therapies: 25(OH)D3 requires the presence of 25-hydroxyvitaminD-1alpha-hydroxylase, whereas 19-nor-1alpha,25(OH)2D2, F6-1alpha,25(OH)2D3 and EB1089 do not. These compounds may be good candidates for human clinical trials in prostate cancer.