Androgen receptor mutations in androgen-independent prostate cancer: Cancer and Leukemia Group B Study 9663

ME Taplin, B Rajeshkumar, S Halabi… - Journal of clinical …, 2003 - ascopubs.org
ME Taplin, B Rajeshkumar, S Halabi, CP Werner, BA Woda, J Picus, W Stadler, DF Hayes
Journal of clinical oncology, 2003ascopubs.org
Purpose: The mechanisms responsible for prostate cancer androgen independence are
diverse. Mutations of the androgen receptor (AR) gene that broaden ligand specificity have
been implicated. Bone marrow specimens containing prostate tumor were obtained from
men undergoing antiandrogen withdrawal for AR sequence analysis and clinical correlation.
Materials and Methods: Eligible men enrolled on a trial of antiandrogen withdrawal had a
minimum prostate-specific antigen (PSA) level of 5 ng/dL that was increasing on castration …
Purpose: The mechanisms responsible for prostate cancer androgen independence are diverse. Mutations of the androgen receptor (AR) gene that broaden ligand specificity have been implicated. Bone marrow specimens containing prostate tumor were obtained from men undergoing antiandrogen withdrawal for AR sequence analysis and clinical correlation.
Materials and Methods: Eligible men enrolled on a trial of antiandrogen withdrawal had a minimum prostate-specific antigen (PSA) level of 5 ng/dL that was increasing on castration therapy including an antiandrogen. With informed consent, marrow biopsies were obtained to collect prostate tumor. Additional samples were obtained from men enrolled on chemotherapy trials. AR cDNA or DNA was polymerase chain reaction–amplified, cloned, and sequenced. The AR CAG repeat length was recorded.
Results: One hundred eighty-four bone marrow biopsies were obtained, and 48 had prostate tumor detected by light microscopy. The ARs from these 48 samples were sequenced. Overall, five (10%) of 48 tumors had mutated ARs. AR point mutations were detected in the hormone-binding domain involved in transcription factor binding. Three mutations were novel in prostate cancer. One tumor sample had a CAG repeat length of 21, compared with germline length of 22 repeats. There was no association between detectability of AR mutations and antiandrogen withdrawal response or survival.
Conclusion: These data suggest that AR mutations are present in approximately 10% of patients with prostate cancer who experience treatment failure with hormone therapy that included an antiandrogen. Mutations in the AR likely confer a growth advantage for a subset of progressive prostate cancers. Correlation of AR mutation with antiandrogen withdrawal response or survival could not be made.
ASCO Publications