Relationship between concentrations of extractable and matrix-bound nuclear androgen receptor and clinical response to endocrine therapy for prostatic adenocarcinoma.

The relationship between androgen receptor concentrations and clinical response to endocrine therapy for prostatic adenocarcinoma was investigated for 13 stage D patients. Both cytoplasmic and nuclear-androgen receptors were quantified. For nuclear androgen receptor, nuclei were isolated and treated with high ionic strength buffer (0.6 M KCl) to yield a KCl-extractable fraction; the nuclei were then treated with DNase I to yield nuclear matrices. Electron microscopy confirmed the relative nuclear purity and revealed matrix morphology. An hydroxylapatite binding assay and methyltrienolone (R1881) were used to quantify androgen receptor in cytosol, the KCl-extract and matrix preparations. Following 6 months of hormonal therapy, the clinical status of patients was re-evaluated and the patients were grouped according to disease response. The androgen receptor data obtained prior to therapy were compared for the disease response groups. The mean concentrations of cytoplasmic androgen receptor, KCl-extractable nuclear androgen receptor and nuclear matrix-bound androgen receptor, respectively, in those patients with disease progression or death (no. = 6), were 671 +/- 232, 45 +/- 17 and 119 +/- 34 fmol. per gm. of tissue +/- S.E.M., and for those with disease regression or stabilization (no. = 7), 1427 +/- 435, 193 +/- 53 and 611 +/- 92 fmol. per gm. of tissue +/- S.E.M. While cytoplasmic androgen receptor concentrations were not related to clinical status, both extractable and matrix-bound nuclear androgen receptor concentrations were significantly higher in the group which responded to hormonal therapy. These results suggest that nuclear-extractable and nonextractable androgen receptor concentrations are useful indices for the prediction of hormone-dependence of prostatic cancer.